Cervical cancer is a primary cause of death in women throughout the world, and early identification using cell classification is essential for increasing survival rates. Timely and accurate detection of cancer cells is critical for promoting personalized treatment and clinical diagnostics. This paper introduces Conv-Jacobian Kolmogorov Arnold Network (Conv-JKAN), a novel deep-learning architecture specifically designed for high-accuracy cervical cancer cell classification from cytological images. The Conv-JKAN uniquely integrates convolutional neural networks with the Jacobian Kolmogorov Arnold network, enabling enhanced feature extraction while maintaining computational efficiency. The model employs just three convolutional layers and three JKAN layers to capture intricate spatial and contextual patterns across varying resolutions, ensuring superior recognition of subtle cytopathological changes. The novelty of Conv-JKAN lies in its hierarchical decomposition approach, where JKAN dissects high-dimensional functions into low-dimensional components using Jacobian polynomial-based sensitivity analysis. This novel technique enhances both interpretability and gradient flow, addressing common challenges in deep-learning-based medical diagnostics. To validate its robustness, Conv-JKAN was rigorously tested on three public cytology datasets, achieving high classification performance: 100

Objective:Colposcopy involves subjective visual assessment of cervical features that may indicate cervical dysplasia. Pattern recognition during colposcopy could be enhanced by artificial intelligence (AI). Using colposcopy images with precisely mapped multiple biopsy sites and corresponding histologic diagnoses, we developed an AI model, Cervix-AID-Net, to classify colposcopy images into low-grade disease [less than cervical intraepithelial neoplasia (CIN) grade 2] and high-grade disease (CIN grade 2 or above). The objective of this study was to compare the diagnostic performance of the Cervix-AID-Net model with the digital colposcope (DySIS) color map and colposcopists' interpretations of the cervix in identifying low-grade and high-grade disease.Methods:The authors used 3,153 colposcopy images from 178 women, each with 4 biopsies, to train and validate the algorithm. Sensitivity, specificity, positive predictive value, negative predictive value, and accuracy were calculated with 95% CIs.Results:Cervix-AID-Net achieved a diagnostic accuracy of 99.8% (95% CI: 99.6-99.9) in classifying colposcopy images into low-grade and high-grade categories. This was significantly higher than the DySIS color map accuracy of 58.8% (95% CI: 51.1-66.1) and the accuracy of the colposcopist's visual impression of the cervix (55.1%, 95% CI: 47.2%-62.5%).Conclusion:This first version of the Cervix-AID-Net demonstrated superior diagnostic accuracy compared with both the DySIS color map and colposcopists' visual assessment. The results need confirmation in a prospective clinical trial.

In this study, we explore the application of deep learning techniques for predicting cleansing quality in colon capsule endoscopy (CCE) images. Using a dataset of 500 images labeled by 14 clinicians on the Leighton-Rex scale (Poor, Fair, Good, and Excellent), a ResNet-18 model was trained for classification, leveraging stratified K-fold cross-validation to ensure robust performance. To optimize the model, structured pruning techniques were applied iteratively, achieving significant sparsity while maintaining high accuracy. Explainability of the pruned model was evaluated using Grad-CAM, Grad-CAM++, Eigen-CAM, Ablation-CAM, and Random-CAM, with the ROAD method employed for consistent evaluation. Our results indicate that for a pruned model, we can achieve a cross-validation accuracy of 88

IntroductionIncreasing demand for colonoscopy continues to strain healthcare systems worldwide. Colon capsule endoscopy (CCE) offers a minimally invasive alternative, but its adoption is limited by high re-investigation rates. The aim of this study is to develop and evaluate clinical prediction models for selecting faecal immunochemical test (FIT) positive patients most suitable for CCE versus colonoscopy.MethodsWe conducted a secondary analysis of data from CareForColon2015 randomized controlled trial (2020-2022), including individuals aged 50-74 years with a positive FIT. Logistic regression models were developed to predict CCE transit, bowel cleansing, completeness, and colonoscopy indication. Sixty candidate predictors were assessed, including demographics, lifestyle factors, FIT values, medications, perceived stress, and health literacy. Models were validated using repeated random subsampling and evaluated on a 10% hold-out set using the area under the receiver-operating-characteristic curve (AUC), Cohen's K, and accuracy. Decision curve analysis (DCA) was performed to assess clinical utility.ResultsCCE achieved complete transit in 92.1% and acceptable bowel cleansing in 71.3% of participants, with 69.6% of investigations deemed complete. Colonoscopy was indicated in 68.0% of cases, based on broad inclusion criteria, and 55.9%, based on more stringent criteria. Models predicting colonoscopy indication showed moderate performance (AUC 0.69-0.71; accuracy 65-67%; Cohen's K 0.28-0.30). DCA indicated positive net benefit for both models within threshold probabilities of 0.5-0.75, supporting their potential to identify FIT-positive patients unlikely to benefit from immediate colonoscopy.ConclusionsClinical prediction models may assist in post-FIT triage between CCE and colonoscopy. DCA suggests potential to reduce unnecessary colonoscopies by identifying low-risk patients suitable for initial CCE. External validation is needed before clinical implementation.

Human cortical organoids provide an experimentally accessible model of early neural circuit formation, yet whether their activity reflects structured information processing rather than spontaneous synchronization is unclear. We developed a graph-computational framework to quantify stimulus-evoked propagation. This includes stimulus-conditioned functional graphs, a graph-constrained dynamical (graph-neural-network) model used as a system-identification tool, a biological message-passing principle bounding integration depth by observable propagation depth, and a suite of graph-level metrics. We carried this program out in full on longitudinal HD-MEA recordings from three organoids. Once the true acquisition sampling rate and stimulus timing were recovered, the evoked response proved to be a fast, near-synchronous network burst with no measurable outward propagation (peak-latency vs. distance slope = 0). The propagation/integration-depth metrics (Deff ,reachability index, dmax) therefore do not apply, and per-day connectivity graphs were not reliably estimable at the available trial count, a negative result with methodological consequences for applying such metrics to organoid data. Reframing around synchrony, response-population size and shared variability revealed a control-validated phenomenon, i.e., repeated daily stimulation progressively depressed and spatially contracted the evoked response. That repeated stimulation reshapes organoid networks is established, but longitudinal designs in which every preparation is stimulated cannot separate this from developmental maturation. We break that confound with a developmentally-matched, stimulation-naive control, where at day 7, an organoid receiving its first-ever stimulation engaged 93

ABSTRACT Depression is a common and devastating mental health illness with serious personal and societal consequences. Despite advancing treatment techniques, there are still hurdles in the effective diagnosis and treatment of depression, such as prompt diagnosis, personalized medication, and continuous monitoring. In recent years, artificial intelligence (AI) has emerged as a potential tool in mental health treatment, providing novel solutions to these difficulties. This systematic study aims to comprehensively assess the existing AI systems for depression detection and treatment. The paper presents a systematic and comprehensive review of the last decade for depression detection, prediction, and treatment. One hundred eighty journal articles fulfilling preset inclusion criteria were found and analyzed using Preferred Reporting Items for Systematic Reviews and Meta‐Analyses from major academic databases. This review used a variety of detection modalities (physical, physiological, repetitive transcranial magnetic stimulation, and pharmacological treatment response) and AI approaches, including machine learning (ML) and deep learning (DL), to address various areas of depression care, including detection, diagnosis, prediction, and treatment. Key findings demonstrate that AI offers tremendous promise in boosting depression care across the continuum, from early identification to individualized therapy optimization and remote monitoring. ML and DL models demonstrate promising accuracy in predicting depression onset, severity, and treatment response based on diverse data sources, including electroencephalogram, electrocardiogram, photoplethysmography, electrodermal activity, electronic healthcare records, facial, speech, text, and pharmaceutical data. The paper highlights the important research challenges in current automated depression decision‐making models. Finally, we emphasize the prospects for developing effective and robust AI‐based depression models incorporating data and model fusion, the model's trust, portability, privacy preservation, and security features. This article is categorized under: Fundamental Concepts of Data and Knowledge > Explainable AI Technologies > Machine Learning Technologies > Artificial Intelligence

Machine unlearning has garnered significant attention due to its ability to selectively erase knowledge obtained from specific training data samples in an already trained machine learning model. This capability enables data holders to adhere strictly to data protection regulations. However, existing unlearning techniques face practical constraints, often causing performance degradation, demanding brief fine-tuning post unlearning, and requiring significant storage. In response, this paper introduces a novel class of machine unlearning algorithms. First method is partial amnesiac unlearning, integration of layer-wise pruning with amnesiac unlearning. In this method, updates made to the model during training are pruned and stored, subsequently used to forget specific data from trained model. The second method assimilates layer-wise partial-updates into label-flipping and optimization-based unlearning to mitigate the adverse effects of data deletion on model efficacy. Through a detailed experimental evaluation, we showcase the effectiveness of proposed unlearning methods. Experimental results highlight that the partial amnesiac unlearning not only preserves model efficacy but also eliminates the necessity for brief post fine-tuning, unlike conventional amnesiac unlearning. Moreover, employing layer-wise partial updates in label-flipping and optimization-based unlearning techniques demonstrates superiority in preserving model efficacy compared to their naive counterparts.

BACKGROUND AND OBJECTIVE:Cervical cancer remains a major worldwide health issue, with high morbidity and mortality rates if diagnosed and treated at a later stage. Early identification and risk assessment are crucial for preventive interventions. METHODS:This paper presents the Cervix-AID-Net model for classifying cervical precancer risk using still images captured from a DYSIS colposcope. The study designs and evaluates the proposed Cervix-AID-Net model to classify high-risk and low-risk cervical precancer classes. The model comprises a Convolutional Block Attention Module (CBAM) and convolutional layers that extract interpretable and representative features from colposcopic images to distinguish high-risk and low-risk cervical precancer. In addition, the proposed Cervix-AID-Net model integrates gradient class activation maps, Local Interpretable Model-agnostic Explanations, CartoonX, and pixel rate distortion techniques to explain model decisions using output feature maps and input features. RESULTS:The evaluation using holdout and ten-fold cross-validation techniques yielded classification accuracies of 99.33% and 99.81%, respectively. The analysis revealed that CartoonX provides meticulous explanations for the decision of the Cervix-AID-Net model due to its ability to provide the relevant piecewise smooth part of the image. The effect of Gaussian noise and blur on the input shows that the performance remains unchanged up to Gaussian noise of 3% and blur of 10%, while the performance decreases thereafter. A comparison study of the proposed model's performance with other deep learning approaches highlights the Cervix-AID-Net model's potential as a supplemental tool for increasing the effectiveness of cervical precancer risk assessment. CONCLUSIONS:The proposed method, which incorporates CBAM and explainable artificial intelligence, has the potential to influence the prevention and early detection of cervical cancer. Thus, the proposed framework will help improve patient outcomes and reduce the worldwide burden of this preventable disease.

Despite recent surge of interest in deploying colon capsule endoscopy (CCE) for early diagnosis of colorectal diseases, there remains a large gap between the current state of CCE in clinical practice, and the state of its counterpart optical colonoscopy (OC). This is due to several factors, such as low quality bowel cleansing, logistical challenges around both delivery and collection of the capsule, and most importantly, the tedious manual assessment of images after retrieval. Our study, built on the "Danish CareForColon2015 trial (cfc2015)" is aimed at closing this gap, by focusing on the full integration of AI in CCE's pathway, where image processing steps linked to the detection, localization and characterisation of important findings are carried out autonomously using various AI algorithms. We developed a family of algorithms based on explainable deep neural networks (DNN) that detect polyps within a sequence of images, feed only those images containing polyps into two parallel independent networks to characterize, and estimate the size of important findings. Our recognition DNN to detect colorectal polyps was trained and validated ([Formula: see text]) and tested ([Formula: see text]) on an unaugmented database of 1751 images containing colorectal polyps and 1672 images of normal mucosa reached an impressive sensitivity of [Formula: see text], a specificity of [Formula: see text], and a negative predictive value (NPV) of [Formula: see text]. The characterisation DNN trained on an unaugmented database of 317 images featuring neoplastic polyps and 162 images of non-neoplastic polyps reached a sensitivity of [Formula: see text] and a specificity of [Formula: see text] in classifying polyps. The size estimation DNN trained on an unaugmented database of 280 images reached an accuracy of [Formula: see text] in correctly segmenting the polyps. By automatically incorporating important information including size, location and pathology of the findings into CCE's pathway, we moved a step closer towards the full integration of explainable AI (XAI) in CCE's routine clinical practice. This translates into a fewer number of unnecessary investigations and resection of diminutive, insignificant colorectal polyps.

Background and Aim: Colon capsule endoscopy (CCE) offers a minimally invasive method for imaging gastrointestinal lesions, including colorectal polyps, which may be precursors to colorectal cancer. However, its low image quality poses challenges for tasks such as polyp characterization. This work develops a low-complexity AI model, ResNet9-KAN, by integrating the Kolmogorov-Arnold network (KAN) into 9-layer residual network (ResNet9) architecture. This model efficiently characterizes polyps as neoplastic or non-neoplastic in CCE images, facilitating real-time patient management. Methods: This work utilized a CCE dataset generated from the PillCam Colon 2 system at four hospitals in the Region of Southern Denmark. It comprises 2089 CCE images of 479 polyps (317 neoplastic, 162 non-neoplastic) from a bowel cancer screening population aged 50 to 74. The proposed ResNet9-KAN and several existing AI models were trained on 1672 CCE images (221 neoplastic, 113 non-neoplastic polyps) and evaluated on 569 test images (48 neoplastic, 25 non-neoplastic polyps). Results: The evaluation revealed that our proposed ResNet9-KAN surpassed existing AI models with per-image characterization accuracy of 97.71 %, demonstrating an excellent balance between sensitivity (97.10 %) and specificity (98.17 %). It also achieved the highest F1 score of 0.9730 and a competitive area under the curve (AUC) of 0.9895. Additionally, ResNet9-KAN exhibited per-polyp characterization accuracy of 99.23 %, with a sensitivity of 99.85 %, specificity of 98.65 %, and an F1 score of 0.9912. Conclusions: This work highlights the efficacy of ResNet9-KAN in accurately characterizing polyps in low-quality CCE images, showing substantial potential for in situ characterization where histological verification currently requires a follow-up colonoscopy.

Early diagnosis of abnormal cervical cells enhances the chance of prompt treatment for cervical cancer (CrC). Artificial intelligence (AI)-assisted decision support systems for detecting abnormal cervical cells are developed because manual identification needs trained healthcare professionals, and can be difficult, time-consuming, and error-prone. The purpose of this study is to present a comprehensive review of AI technologies used for detecting cervical pre-cancerous lesions and cancer. The review study includes studies where AI was applied to Pap Smear test (cytological test), colposcopy, sociodemographic data and other risk factors, histopathological analyses, magnetic resonance imaging-, computed tomography-, and positron emission tomography-scan-based imaging modalities. We performed searches on Web of Science, Medline, Scopus, and Inspec. The preferred reporting items for systematic reviews and meta-analysis guidelines were used to search, screen, and analyze the articles. The primary search resulted in identifying 9745 articles. We followed strict inclusion and exclusion criteria, which include search windows of the last decade, journal articles, and machine/deep learning-based methods. A total of 58 studies have been included in the review for further analysis after identification, screening, and eligibility evaluation. Our review analysis shows that deep learning models are preferred for imaging techniques, whereas machine learning-based models are preferred for sociodemographic data. The analysis shows that convolutional neural network-based features yielded representative characteristics for detecting pre-cancerous lesions and CrC. The review analysis also highlights the need for generating new and easily accessible diverse datasets to develop versatile models for CrC detection. Our review study shows the need for model explainability and uncertainty quantification to increase the trust of clinicians and stakeholders in the decision-making of automated CrC detection models. Our review suggests that data privacy concerns and adaptability are crucial for deployment hence, federated learning and meta-learning should also be explored. This article is categorized under: Fundamental Concepts of Data and Knowledge > Explainable AI Technologies > Machine Learning Technologies > Classification

Emotion recognition is the ability to precisely infer human emotions from numerous sources and modalities using questionnaires, physical signals, and physiological signals. Recently, emotion recognition has gained attention because of its diverse application areas, like affective computing, healthcare, human-robot interactions, and market research. This paper provides a comprehensive and systematic review of emotion recognition techniques of the current decade. The paper includes emotion recognition using physical and physiological signals. Physical signals involve speech and facial expression, while physiological signals include electroencephalogram, electrocardiogram, galvanic skin response, and eye tracking. The paper provides an introduction to various emotion models, stimuli used for emotion elicitation, and the background of existing automated emotion recognition systems. This paper covers comprehensive searching and scanning of wellknown datasets followed by design criteria for review. After a thorough analysis and discussion, we selected 142 journal articles using PRISMA guidelines. The review provides a detailed analysis of existing studies and available datasets of emotion recognition. Our review analysis also presented potential challenges in the existing literature and directions for future research.

Electronic health records (EHR) of large populations constitute a vast untapped resource for data-driven diagnosis and disease progression. We develop a model capable of predicting future steps in a patient’s journey for prostate cancer (PC) and its metastases without relying on direct biomarker-measurements on a set of $18\,529$ EHR. To this end, we 1) harmonise EHR without presumptions–events are sorted and grouped by fundamental a priori principles; 2) develop a new Long-Short-Term Memory (LSTM) recurrent neural network node for learning temporal relations, on which we build an autoencoder based model; 3) derive a graph representation based on unsupervised $k$ -means clustering of events related to PC in the autoencoder’s latent layer. We report $88 {\%}$ predicting accuracy for the targeted metastasis-related events, and lower accuracies for more general events. The model gains interpretability with a graph representation illustrating the patient journey. Most importantly, we predict that $20 {\%}$ of all PC diagnosed patients will progress into metastatic disease one visit ahead of time. For the remaining patients we can predict the next step in their journey. We conclude that the model based on the new LSTM node provides a valuable tool for earlier diagnosis of life threatening metastases and quality assurance of the procedure.

ObjectiveIdentification of patients at high risk of aggressive prostate cancer is a major clinical challenge. With the view of developing artificial intelligence-based methods for identification of these patients, we are constructing a comprehensive clinical database including 7448 prostate cancer (PCa) Danish patients. In this paper we provide an epidemiological description and patients' trajectories of this retrospective observational population, to contribute to the understanding of the characteristics and pathways of PCa patients in Denmark.ResultsIndividuals receiving a PCa diagnosis during 2008-2014 in Region Southern Denmark were identified, and all diagnoses, operations, investigations, and biochemistry analyses, from 4 years prior, to 5 years after PCa diagnosis were obtained. About 85.1% were not diagnosed with metastatic PCa during the study period (unaggressive PCa); 9.2% were simultaneously diagnosed with PCa and metastasis (aggressive-advanced PCa), while 5.7% were not diagnosed with metastatic PCa at first, but they were diagnosed with metastasis at some point during the 5 years follow-up (aggressive-not advanced PCa). Patients with unaggressive PCa had more clinical investigations directly related to PCa detection (prostate ultrasounds and biopsies) during the 4 years prior to PCa diagnosis, compared to patients with aggressive PCa, which may have contributed to the early detection of PCa.

Background and Aims:Colon capsule endoscopy (CCE) faces substantial challenges, one of which is achieving adequate colon cleansing. Furthermore, the interobserver agreement on bowel-cleansing quality varies. To address this issue, we developed an artificial intelligence algorithm (AIA) to evaluate bowel-cleansing quality. The aim of this study was to estimate the interobserver agreement on bowel cleansing between a group of experienced CCE readers and an AIA and to examine whether percentiles of the overall bowel-cleansing quality are a suitable way of reporting the results generated by the AIA. Methods:Bowel-cleansing quality in 842 CCE investigations was scored on both a 2- and 4-point grading scale for the entire colon and by segment by experienced CCE readers and the AIA. For the algorithm, a score was given based on the mean score, median, upper and lower quartiles, and second and 98th percentiles. The level of agreement was evaluated using Cohen's κ. Results:The interobserver agreement between the CCE readers and AIA on bowel-cleansing quality was minimal to none for the overall bowel evaluation, by segment, and on the 2- and 4- point grading scale regardless of the threshold for the AIA score. Conclusions:We found minimal agreement on evaluation of bowel-cleansing quality in CCE between CCE readers and the AIA. Mean or percentiles of the AIA grading did not seem suitable for AI-generated bowel-cleansing evaluation.

Colon capsule endoscopy (CCE) as a novel 2D biomedical image modality based on visible light provides a higher perspective of the potential gastrointestinal lesions like polyps within the small and large intestines than the conventional colonoscopy. As the quality of images acquired via CCE imagery is low, so the artificial intelligence methods are proposed to help detect and localize polyps within an acceptable level of efficiency and performance. In this paper, a new deep neural network architecture known as AID-U-Net is proposed. AID-U-Net consists of two distinct types of paths: a) Two main contracting/expansive paths, and b) Two sub-contracting/expansive paths. The playing role of the main paths is to localize polyps as the target objectives in high resolution and multi-scale manner, while the two sub paths are responsible for preserving and conveying the information of low resolution and low-scale target objects. Furthermore, the proposed network architecture provides simplicity so that the model can be deployed for real time processing. AID-U-Net with an implementation of a VGG19 backbone shows better performance to detect polyps in CCE images in comparison with the other state-of-the-art U-Net models like conventional U-Net, U-Net++, and U-Net3+ with different pre-trained backbones like ImageNet, VGG19, ResNeXt50, Resnet50, InceptionV3 and InceptionResNetV2.

For years, hepatologists have been seeking non-invasive methods able to detect significant liver fibrosis. However, no previous algorithm using routine blood markers has proven to be clinically appropriate in primary care. We present a novel approach based on artificial intelligence, able to predict significant liver fibrosis in low-prevalence populations using routinely available patient data. We built six ensemble learning models (LiverAID) with different complexities using a prospective screening cohort of 3352 asymptomatic subjects. 463 patients were at a significant risk that justified performing a liver biopsy. Using an unseen hold-out dataset, we conducted a head-to-head comparison with conventional methods: standard blood-based indices (FIB-4, Forns and APRI) and transient elastography (TE). LiverAID models appropriately identified patients with significant liver stiffness (> 8 kPa) (AUC of 0.86, 0.89, 0.91, 0.92, 0.92 and 0.94, and NPV ≥ 0.98), and had a significantly superior discriminative ability ( p < 0.01) than conventional blood-based indices (AUC = 0.60–0.76). Compared to TE, LiverAID models showed a good ability to rule out significant biopsy-assessed fibrosis stages. Given the ready availability of the required data and the relatively high performance, our artificial intelligence-based models are valuable screening tools that could be used clinically for early identification of patients with asymptomatic chronic liver diseases in primary care.

Wireless capsule endoscopy (WCE) is a modern, non-invasive method of gastrointestinal examination that can significantly reduce mortality and morbidity. One of the current challenges in WCE is the precise localization of the capsule. An accurate path loss propagation model can be used to find the exact distance from the surface to the capsule inside the abdominal cavity. Unfortunately, there are no standardized In-to-On-Body channel models describing the signal propagation at ultra-high frequencies that are used in the most commercially available WCE systems. This study addresses the gap by conducting an experimental validation of a new propagation model for WCE applications at 2.45 GHz. The results were confirmed by conducting two separate in-vivo trials on porcine animals under general anesthesia. The performance of the model as well as the corresponding ranging errors were evaluated when it was used as an inverse solution for distance estimation to an ingested transmitter. The main advantage of the model is its theoretical basis, which can help further generalize the findings for similar communication scenarios. The obtained ranging error was smaller than one centimeter, suggesting that it can be used for accurate range-based positioning of implanted transmitters.

Semantic segmentation of biomedical images found its niche in screening and diagnostic applications. Recent methods based on deep learning convolutional neural networks have been very effective, since they are readily adaptive to biomedical applications and outperform other competitive segmentation methods. Inspired by the U-Net, we designed a deep learning network with an innovative architecture, hereafter referred to as AID-U-Net. Our network consists of direct contracting and expansive paths, as well as a distinguishing feature of containing sub-contracting and sub-expansive paths. The implementation results on seven totally different databases of medical images demonstrated that our proposed network outperforms the state-of-the-art solutions with no specific pre-trained backbones for both 2D and 3D biomedical image segmentation tasks. Furthermore, we showed that AID-U-Net dramatically reduces time inference and computational complexity in terms of the number of learnable parameters. The results further show that the proposed AID-U-Net can segment different medical objects, achieving an improved 2D F1-score and 3D mean BF-score of 3.82% and 2.99%, respectively.

Gastrointestinal (GI) tract diseases are responsible for substantial morbidity and mortality worldwide, including colorectal cancer, which has shown a rising incidence among adults younger than 50. Although this could be alleviated by regular screening, only a small percentage of those at risk are screened comprehensively, due to shortcomings in accuracy and patient acceptance. To address these challenges, we designed an artificial intelligence (AI)-empowered wireless video endoscopic capsule that surpasses the performance of the existing solutions by featuring, among others: (1) real-time image processing using onboard deep neural networks (DNN), (2) enhanced visualization of the mucous layer by deploying both white-light and narrow-band imaging, (3) on-the-go task modification and DNN update using over-the-air-programming and (4) bi-directional communication with patient’s personal electronic devices to report important findings. We tested our solution in an in vivo setting, by administrating our endoscopic capsule to a pig under general anesthesia. All novel features, successfully implemented on a single platform, were validated. Our study lays the groundwork for clinically implementing a new generation of endoscopic capsules, which will significantly improve early diagnosis of upper and lower GI tract diseases.