INTRODUCTION:Large language models, in particular ChatGPT, have showcased remarkable language processing capabilities. Given the substantial workload of university medical staff, this study aims to assess the quality of multiple-choice questions (MCQs) produced by ChatGPT for use in graduate medical examinations, compared to questions written by university professoriate staffs based on standard medical textbooks. METHODS:50 MCQs were generated by ChatGPT with reference to two standard undergraduate medical textbooks (Harrison's, and Bailey & Love's). Another 50 MCQs were drafted by two university professoriate staff using the same medical textbooks. All 100 MCQ were individually numbered, randomized and sent to five independent international assessors for MCQ quality assessment using a standardized assessment score on five assessment domains, namely, appropriateness of the question, clarity and specificity, relevance, discriminative power of alternatives, and suitability for medical graduate examination. RESULTS:The total time required for ChatGPT to create the 50 questions was 20 minutes 25 seconds, while it took two human examiners a total of 211 minutes 33 seconds to draft the 50 questions. When a comparison of the mean score was made between the questions constructed by A.I. with those drafted by humans, only in the relevance domain that the A.I. was inferior to humans (A.I.: 7.56 +/- 0.94 vs human: 7.88 +/- 0.52; p = 0.04). There was no significant difference in question quality between questions drafted by A.I. versus humans, in the total assessment score as well as in other domains. Questions generated by A.I. yielded a wider range of scores, while those created by humans were consistent and within a narrower range. CONCLUSION:ChatGPT has the potential to generate comparable-quality MCQs for medical graduate examinations within a significantly shorter time.

BACKGROUND:The chatbot application Bennie and the Chats was introduced due to the outbreak of COVID-19, which is aimed to provide substitution for teaching conventional clinical history-taking skills. It was implemented with DialogFlow with preset responses, which consists of a large constraint on responding to different conversations. The rapid advancement of artificial intelligence, such as the recent introduction of ChatGPT, offers innovative conversational experiences with computer-generated responses. It provides an idea to develop the second generation of Bennie and the Chats. As the epidemic slows, it can become an assisting tool for students as additional exercise. In this work, we present the second generation of Bennie and the Chats with ChatGPT, which provides room for flexible and expandable improvement. METHODS:The objective of this research is to examine the influence of the newly proposed chatbot on learning efficacy and experiences in bedside teaching, and its potential contributions to international teaching collaboration. This study employs a mixed-method design that incorporates both quantitative and qualitative approaches. From the quantitative approach, we launched the world's first cross-territory virtual bedside teaching with our proposed application and conducted a survey between the University of Hong Kong (HKU) and the National University of Singapore (NUS). Descriptive statistics and Spearman's Correlation were applied for data analysis. From the qualitative approach, a comparative analysis was conducted between the two versions of the chatbot. And, we discuss the interrelationship between the quantitative and qualitative results. RESULTS:For the quantitative result, we collected a questionnaire from 45 students about the evaluation of virtual bedside teaching between territories. Over 75% of the students agreed that teaching can enhance learning effectiveness and experience. Moreover, by exchanging patients cases, 82.2% of students agreed that it helps to gain more experiences with diseases that may not be prevalent in their own locality. For the qualitative result, the new chatbot provides better usability and flexibility. CONCLUSION:Virtual bedside teaching with chatbots has revolutionized conventional bedside teaching by its advantages and allowing international collaborations. We believe that the training of history taking skills by chatbot will be a feasible supplementary teaching tool to conventional bedside teaching.

Background and Objective:With an increasing number of non-palpable breast lesions detected due to improved screening, accurate localization of these lesions for surgery is crucial. This literature review explores the evolution of localization methods for non-palpable breast lesions, highlighting the translational journey from concept to clinical practice. Methods:A comprehensive search of PubMed, Embase, and Scopus databases until September 2023 was conducted. Key Content and Findings:Multiple methods have been developed throughout the past few decades. (I) Wire-guided localization (WGL) introduced in 1966, has become a reliable method for localization. Its simplicity and cost-effectiveness are its key advantages, but challenges include logistical constraints, patient discomfort, and potential wire migration. (II) Intraoperative ultrasound localization (IOUS) has shown promise in ensuring complete lesion removal with higher negative margin rates. However, its utility is limited to lesions visible on ultrasound (US) imaging. (III) Breast biopsy marker localization: the use of markers has improved the precision of localization without the need for wire. However, marker visibility remains a challenge despite improvements in their design. (IV) Radioactive techniques: radio-guided occult lesion localization (ROLL) and radioactive seed localization (RSL) offer flexibility in scheduling and improved patient comfort. However, they require close multidisciplinary collaboration and specific equipment due to radioactive concerns. (V) Other wireless non-radioactive techniques: wireless non-radioactive techniques have been developed in recent three decades to provide flexible and patient-friendly alternatives. It includes magnetic seed localization, radar techniques, and radiofrequency techniques. Their usage has been gaining popularity due to their safety profile and allowance of more flexible scheduling. However, their high cost and need for additional training remain a barrier to a wider adoption. Conclusions:The evolution of breast lesion localization methods has progressed to more patient-friendly techniques, each with its unique advantages and limitations. Future research on patient-reported outcomes, cosmetic outcomes, breast biopsy markers and integration of augmented reality with breast lesion localization are needed.

Breast cancer is the most common cancer among women globally and can be classified according to various histological subtypes. Current treatment strategies are typically based on the cancer stage and molecular subtypes. This article aims to address the knowledge gap in the understanding of rare breast cancer. A retrospective study was conducted on 4393 breast cancer patients diagnosed from 1992 to 2012, focusing on five rare subtypes: mucinous, invasive lobular, papillary, mixed invasive and lobular, and pure tubular/cribriform carcinomas. Our analysis, supplemented by a literature review, compared patient characteristics, disease characteristics, and survival outcomes of rare breast cancer patients with invasive carcinoma (not otherwise specified (NOS)). Comparative analysis revealed no significant difference in overall survival rates between these rare cancers and the more common invasive carcinoma (NOS). However, mucinous, papillary, and tubular/cribriform carcinomas demonstrated better disease-specific survival. These subtypes presented with similar characteristics such as early detection, less nodal involvement, more hormonal receptor positivity, and less human epidermal growth factor receptor 2 (HER2) positivity. To conclude, our study demonstrated the diversity in the characteristics and prognosis of rare breast cancer histotypes. Future research should be carried out to investigate histotype-specific management and targeted therapies, given their distinct behavior.

Background; The application of artificial intelligence (AI) like Large Language Models (LLM) into the healthcare system has been a frequently discussed topic in recent years. Materials and Methods; We conducted a systemic review on primary studies about the applications of LLM in breast conditions. The studies are then categorized into their respective domains, namely diagnosis, management recommendations and communication for patients. Results; The diagnostic accuracy ranged from 74.3% to 99.6% across different investigation modalities. The concordance of management recommendations ranged from 50% to 70% while the prognostic evaluation of breast cancer patients of distant recurrence showed an accuracy of 75% to 88%. In regards to patient communication, it is revealed that 18-30% of the references used by the LLM were irrelevant. Conclusion; This study highlights the potential benefits of LLM in strengthening patient communication, diagnose and management of patients with breast conditions. With standardized protocol and guideline to minimize potential risks, LLM can be a valuable tool to support future clinicians in the field of breast management. ### Competing Interest Statement The authors have declared no competing interest. ### Funding Statement This study did not receive any funding ### Author Declarations I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained. Yes The details of the IRB/oversight body that provided approval or exemption for the research described are given below: The study used ONLY openly available human data that were originally located at Pubmed, SCOPUS, and Google Scholar databases I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals. Yes I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance). Yes I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable. Yes All data produced in the present work are contained in the manuscript

Background and Objective: Nipple sparing mastectomy (NSM) has become increasingly popular in recent years, both as a treatment for selected cases of breast cancer and for BRCA carriers in prophylactic and therapeutic use. Methods: PubMed, EMBASE, CINAHL and Cochrane were searched for studies on NSM up till June 2020. Key Content and Findings: Different incisions are indicated in different situations and the ultimate decision should be reached after a multi -disciplinary discussion. A common complication of NSM is nipple areolar complex (NAC) necrosis. Therefore, it is imperative for its blood supply to be preserved as much as possible. A deeper understanding of the breast vasculature would also be useful in planning future surgical approach. While NSM offers a superb cosmetic outcome, oncological safety should never be compromised. Short term oncological safety for NSM is evident in recently conducted studies. Three major parameters were evaluated including overall survival, disease -free survival (DFS) and local recurrence rates. In general, NSM is associated with high DFS and low recurrence. As such, carefully selected breast cancer patients and high -risk BRCA patients would benefit from NSM despite occult NAC malignancy being a concern. Conclusions: NSM offers a non -inferior outcome to the conventional mastectomy in selected cases. In the future, randomised controlled trials (RCTs) would be useful in demonstrating the long-term oncologic safety of NSM in breast cancer and BRCA-carrier patients.

PURPOSE:This study aims to evaluate the association between surgical margin status and local recurrence of DCIS. METHODS:A retrospective analysis of a prospectively maintained 20-year DCIS database was performed. >=2 mm margin was defined as clear margin. Local relapse rate between the patients with clear versus close margins were analyzed with Kaplan-Meier analyses. RESULTS:654 patients were analyzed. Median age was 46.5 (Range 18 - 80). 205 (31.3%) were high grade, 194 (29.7%) were intermediate grade, 143 (21.9%) were low grade. 112 (18.3%) were unknown. 202 (30.9%) were estrogen receptor positive, 49 (7.4%) were negative, 403 (61.6%) patients were unknown. 403 (61.6%) patients received mastectomy while 251 (38.4%) patients received BCS and radiotherapy. 549 (83.9%) patients had clear surgical margin, 50 (7.7%) patients had involved (positive) resection margin, 55 (8.4%) had close margin (<2 mm margin). All patients with involved margin received re-excision of margin, while 21 patients (out of 55 who had close resection margins) received re-excision of margin. Negative surgical margins were achieved after the re-excision. 34 patients with close resection margin decided not to receive re-excision but to undergo adjuvant radiotherapy. After median follow-up of 128 months, the 10-year ipsilateral breast tumor relapse (IBTR) was 4.5% (N = 28), Of which 27 (96.4%) patients had clear margin after the initial surgical treatment of DCIS. 1 (3.6%) patient had close surgical margin. Difference in IBTR between the two groups was not statistically significant (p = 0.692). CONCLUSION:Close surgical margin for DCIS is not associated with increased risk of IBTR.

Background This review aims to evaluate the effectiveness of extended reality-based training in surgical education. Methods This systematic review was conducted in line with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. Results A total of 33 studies were included in the qualitative analysis. Nine studies evaluated the effectiveness of virtual reality-based training against no substitutional training. Seven studies looked at training for laparoscopic surgery, and the results were contradicting. Two studies focused on orthopedics training, and the outcomes were positive. Fourteen studies compared the outcomes of virtual reality-based training to conventional didactic teaching, all demonstrating superior outcomes for virtual reality-based training. Nine studies compared the outcomes of virtual reality simulation training to dry lab simulation training. The inferior outcomes of virtual reality simulation training were demonstrated by 5 studies for laparoscopic surgery, 1 study for arthroscopic procedures, 1 study for robotic surgery, and 1 study for dental procedures. One study found potential benefits of virtual reality simulation training on orthopedics surgeries. One study found virtual reality simulation training to be superior to cadaveric training, and 3 studies found augmented reality and virtual reality-based training to be comparable to supervised operative opportunities. Conclusion Extended reality-based training is a potentially useful modality to serve as an adjunct to the current physical surgical training.

A systematic review was conducted in line with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses protocol to evaluate the diagnostic accuracy of artificial intelligence (AI) in ductal carcinoma in situ. Four databases were searched for articles up to December 2022: Embase, PubMed, Scopus, and Web of Science. 23 studies were included, and a search of grey literature was not performed. The following parameters were extracted: the accuracy, sensitivity, specificity, positive predictive value, and negative predictive value of each study. Statistical analysis of the included studies revealed that AI-assisted histopathological analysis is of high accuracy (83.78%), sensitivity (83.88%), and specificity (85.49%) and has a high positive predictive value (89.43%). Our results also reported that convolutional neural network (CNN) is the most commonly used mode of machine learning—21 models used only CNN, whereas 2 models used only support vector machines (SVM). On an average, CNN reported slightly higher accuracy and sensitivity (86.71% and 85.22%, respectively) than SVM (accuracy, 85.00%; sensitivity, 70.00%). When the 2 methods were combined, a mean accuracy of 82.52% and a mean sensitivity of 83.00% were achieved. The use of AI as a diagnostic adjunct can markedly improve the accuracy and efficiency of DCIS diagnosis and can, therefore, reduce pathologists’ workload.

The global pandemic of COVID-19 has led to extensive practice of online learning. Our main objective is to compare different online synchronous interactive learning activities to evaluate students’ perceptions. Moreover, we also aim to identify factors influencing their perceptions in these classes. A cross-sectional, questionnaire-based study focusing on clinical year medical students’ perceptions and feedback was conducted between February 2021 –June 2021 at the University of Hong Kong. Online learning activities were divided into bedside teaching, practical skill session, problem-based learning (PBL) or tutorial, and lecture. A questionnaire based on the Dundee Ready Education Environment Measure (DREEM) was distributed to 716 clinical year students to document their perceptions. One hundred responses were received with a response rate of 15.4% (110/716, including 96 from bedside teaching, 67 from practical skill session, 104 from PBL/tutorial, and 101 from lecture). For the mean score of the DREEM-extracted questionnaire, online PBL/tutorial scored the highest (2.72 ± 0.54), while bedside scored the lowest (2.38 ± 0.68, p = 0.001). Meanwhile, there was no significant difference when we compared different school years (p = 0.39), age (p = 0.37), gender (p = 1.00), year of internet experience (<17 vs ≥17 years p = 0.59), or prior online class experience (p = 0.62). When asked about students’ preference for online vs face-to-face classes. Students showed higher preferences for online PBL/tutorial (2.06 ± 0.75) and lectures (2.27 ± 0.81). Distraction remains a significant problem across all four learning activities. A multivariate analysis was performed regarding students’ reported behavior in comparison with their perception through the DREEM-extracted questionnaire. The results showed that good audio and video quality had a significant and positive correlation with their perception of online bedside teaching, practical skill sessions, and PBL/tutorial. It also showed that the use of the video camera correlated with an increase in perception scores for lectures. The present analysis has demonstrated that students’ perception of different online synchronous interactive learning activities varies. Further investigations are required on minimizing distraction during online classes.

INTRODUCTION:This is a prospective single arm clinical trial on cryosurgery for early breast cancers, to evaluate the expanded criteria to tumors larger than 1.5 cm and non-luminal breast cancers. METHODS:Inclusion criteria include Solitary T1 breast cancers of any immunohistotypes. Cryosurgery was performed using the IceCure ProSense Cryoablation System. Lumpectomy of the cryoablated tumor was then performed 8 weeks after cryosurgery. RESULTS:Fifteen patients underwent cryosurgery followed by lumpectomy (BCS). Median age was 53 years old 5 (33.3%) patients had ductal carcinoma in situ (DCIS), while 10 (66.7%) patients had invasive ductal carcinoma (IDC), of which 5 (50%) patients had luminal type cancers of which 3 (60%) were luminal A and 2 (40%) luminal B, 3 (30%) patients had HER2 enriched invasive carcinoma and 2 (20%) patients had triple negative IDC. Median tumor size was 13mm (Range 8.6-18mm). Seven (46.7%) patients were found to have residual cancer in the post-cryosurgery lumpectomy specimen. All residual cancers were found at the periphery of the cryoablated breast tissue. All breast cancers were otherwise completely ablated centrally as confirmed by routine histopathology, immunochemistry and TUNEL assay for evaluation of cell viability. None of the tumor factors such as tumor biology, as well as surgical factors such as ablation time and iceball size, were associated with risk of residual cancer. None of the 15 patients developed post-operative complications. CONCLUSION:Residual cancer occurs at the periphery of the cryoablation site, careful pre-operative planning and intra-operative monitoring is crucial to ensure complete cryoablation.

BACKGROUND:The incidence of pregnancy-associated breast cancer (PABC) is increasing. Its tumor characteristics and overall survival compared with those in nonpregnant patients remain controversial. While there have been suggestions that PABC patients have a 40 % increase in the risk of death compared to non-pregnant patients, other studies suggested similar disease outcomes. This study aims to review our local experience with PABC. METHODS:Twenty-eight patients diagnosed with PABC and twenty-eight patients diagnosed at premenopausal age randomly selected by a computer-generated system during the same period were recruited. Background characteristics, tumor features, and survival were compared. RESULTS:Among the twenty-eight pregnant patients, seventeen were diagnosed during pregnancy, and eleven were diagnosed in the postpartum period. Compared to the non-pregnant breast cancer patients, they presented with less progesterone receptor-positive tumor (35.7 % vs. 64.2 %, p = 0.03). Although there was no statistically significant difference in tumor size (p = 0.44) and nodal status (p = 0.16), the tumor tended to be larger in size (2.94 +/- 1.82 vs 2.40 +/- 1.69 cm) and with more nodal involvement (35.7 % vs 25.0 %). There was also a trend of delayed presentation to medical attention, with a mean duration of 13.1 weeks in the PABC group and 8.6 weeks in the control group. However, the overall survival did not differ (p = 0.63). CONCLUSION:PABC is increasing in incidence. They tend to have more aggressive features, but overall survival remains similar. A multidisciplinary approach is beneficial for providing the most appropriate care.

ABSTRACTIntroductionThis is a prospective study on the quality of multiple-choice questions (MCQs) generated by the language model ChatGPT for the use in medical graduate examination.Methods50 MCQs were generated by ChatGPT with reference to two standard undergraduate medical textbooks (Harrison’s, and Bailey & Love’s). Another 50 MCQs were drafted by two university professoriate staffs using the same medical textbooks. All 100 MCQ were individually numbered, randomized and sent to five independent international assessors for MCQ quality assessment using a standardized assessment score on five assessment domains; namely, appropriateness of the question, clarity and specificity, relevance, discriminative power of alternatives, and suitability for medical graduate examination.ResultsThe total time required for ChatGPT to create the 50 questions was 20 minutes 25 seconds while it took two human examiners a total of 211 minutes 33 seconds for drafting the 50 questions.When a comparison of the mean score was made between the questions constructed by AI with those drafted by human, only in the relevance domain that the AI was inferior to human (AI: 7.56 +/- 0.94 vs human: 7.88 +/- 0.52; p = 0.04). There was no significant difference in question quality between questions drafted by AI versus human, in the total assessment score as well as in other domains.Questions generated by AI yielded a wider range of scores while those created by human were consistent and within a narrower range.ConclusionChatGPT has the potential to generate comparable-quality MCQs for medical graduate examination within a significantly shorter time.

BACKGROUND:The effect of playing background music on surgical outcomes has been controversial. This prospective case-control study aims to evaluate the impact of music tempo in general surgical settings. STUDY DESIGN:Six hundred consecutive patients with nonmetastatic breast cancer receiving breast cancer surgery have been recruited since April 2017. Patients were then assigned to 3 arms in consecutive order. The surgeon operated without music in study arm A; the surgeon operated with slow music in study arm B; and in study arm C, the surgeon operated with fast background music. Patients' clinical records were reviewed by an independent blinded assessor. RESULTS:Baseline demographic data were comparable among the 3 study arms. Seven (3.5%) patients from study arm A developed minor complications (Clavien-Dindo class I and II); none developed major complications (Clavien-Dindo class III or above). Six (3.0%) patients from study arm B and C, respectively (slow/fast music groups), developed minor complications; none developed a major complication. Mean blood loss was also similar among the 3 study arms (5.1, 5.1, and 5.2 mL, respectively; p > 0.05). Operating time was significantly shorter in study arm C: 115 minutes (90-145), compared with 125 minutes (100-160) in study arm A (p < 0.0001) and 120 minutes (95-155) in study arm B (p = 0.0024). After a median follow-up of 40 months (3-56), 40 months (3-56), and 39.5 months (3-56), the local recurrence rates were 1.5%, 1%, and 1%, respectively (p > 0.05). CONCLUSION:Playing music in the operating room is safe in general surgical settings in experienced hands.

Journal of the American College of Surgeons 236(1):p 275-276, January 2023. | DOI: 10.1097/XCS.0000000000000427

Introduction: A novel chatbot mobile app for training of undergraduate medical students' clinical history taking skills was developed in 2021. Students were able to take clinical history from the virtual patient for bedside teaching. A case-control study was conducted to evaluate the effectiveness of learning with chatbot mobile app, versus conventional bedside teachings with real patients. Methods: 132 final year medical students were randomized into two groups - Conventional bedside teaching with clinical history taken from a real patient, and Bedside teaching with clinical history taken from the Chatbot. Independent blinded assessment of students' history taking skills was conducted. Students' performance were assessed by standardized marking scheme. Results: Median age was 23 years old (Range 21-30 years old). There were 62 female and 70 male students. 64 students were randomized into conventional group while 68 students were randomized into the chatbot group. Baseline demographic data were comparable between the two groups. Blinded assessment of students' performance in clinical history taking were comparable between the conventional group and chatbot group (p > 0.05). Conclusion: With the promising results we have demonstrated in this study, we believe training of history taking skills by chatbot will be a feasible alternative to conventional bedside teaching.