MKMustafa Tamim Alam Khan

Mustafa Tamim Alam Khan

*This is an algorithmically generated author record based on publicly available information.

Mustafa Tamim Alam Khan has a total of 98 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    AI through the looking glass: a controlled feasibility study of wearable AI assistance on plastic surgery in-service examination questions

    by Mustafa Khan, Joao Bombardelli, Mahwish Khan, Erik Marques, Mohin Bhadkamkar

    Purpose: Large language models (LLMs) have demonstrated substantial accuracy on specialty in-service examinations. Commercially available smart glasses now make real-time AI assistance wearable, raising concerns about testing integrity. We developed and evaluated a smart glasses application that captures an examination question, submits it to an AI model, and delivers the answer directly to the wearer, to assess the feasibility and limitations of this workflow in a controlled simulation. Methods: In Phase 1, three LLMs (Google Gemini 2.5 Flash, OpenAI GPT-4o, and Anthropic Claude Sonnet 4.6) and a majority-vote ensemble were benchmarked on 743 Plastic Surgery In-Service Training Examination (PSITE) question texts from the 2021–2023 examination cycles. Clinical images were not supplied. In Phase 2, the model with the highest observed accuracy was deployed through Rokid AI Glasses, a companion phone, and a cloud service. One study author tested the system on a 50-question subset selected from Phase 1. Outcomes were accuracy, first-attempt success, eventual completion, and successful-attempt latency. Results: Gemini 2.5 Flash achieved the highest individual accuracy (82.8%), followed by Claude Sonnet 4.6 (80.9%) and GPT-4o (76.9%). The majority-vote ensemble achieved 81.6% accuracy and did not outperform Gemini. In Phase 2, the wearable system answered 76.0% of questions correctly (38/50; 95% CI, 62.6%–85.7%). Answers were returned on the first attempt for 45/50 questions (90.0%); all 50 were completed after 56 attempts. Six attempts failed because of application timeouts requiring restart. Mean successful-attempt latency was 8.9 s (SD 4.2), excluding failed attempts and restart time. Conclusions: Commercially available smart glasses and AI services can be connected into a workflow that answers plastic surgery examination questions in real time, although errors and retries limit reliability. This controlled simulation illustrates a potential route to unauthorized assistance. The findings aim to encourage broader discussion of emerging risks to testing integrity and the future of education and assessment.

  • Embodied AI in the operating room: a voice-interactive robotic scrub nurse for surgical instrument handoffs

    by Mustafa Khan, Joao Bombardelli

    Embodied artificial intelligence may extend surgical robotics beyond teleoperation by enabling robots to interpret natural-language commands, perceive dynamic environments, and generate adaptive physical actions. Surgical instrument exchange provides a clinically recognizable test case requiring communication, visual localization, dexterity, and bidirectional interaction. We developed a voice-interactive robotic scrub nurse through task-specific adaptation of the pretrained π0.5 vision-language-action model capable of voice-conditioned surgical instrument handoff, take-back, and spoken acknowledgment. A bimanual robotic platform was trained by fine-tuning the pretrained π0.5 model on 350 human-teleoperated demonstrations. The system integrated real-time multicamera observations, Whisper large-v2 speech recognition, language-conditioned robotic action generation, and text-to-speech acknowledgment. Performance was evaluated in 100 autonomous trials across five tasks: needle-driver handoff, forceps handoff, needle-and-suture handoff, needle-driver take-back, and forceps take-back. Each task included 10 trials using spatial configurations represented during training and 10 trials using held-out spatial configurations within the same environment. Outcomes included command interpretation, stage-wise completion, end-to-end success, completion time, and failure mode. Recovery after interference was examined qualitatively in separate representative demonstrations. The system completed 81% of trials successfully. Success was 95% for both take-back tasks, 85% for forceps handoff, 80% for needle-driver handoff, and 50% for needle-and-suture handoff. Performance decreased from 92% in seen layouts to 70% in held-out layouts. Speech-command interpretation succeeded in 99% of trials. Once a secure grasp was achieved, downstream manipulation succeeded in 81 of 82 trials. Missed grasp was the predominant failure mode. A unified learned policy enabled voice-conditioned bidirectional surgical instrument exchange and demonstrated meaningful generalization to held-out layouts within the same environment. Improved grasp acquisition, expanded instrument diversity, quantitative recovery testing, and evaluation in realistic operating-room workflows are required before clinical translation or integration with complete back-table workflows.


  • Open Access

    Ligand-receptor interactions induce and mediate regulatory functions of BATF3 + B cells

    by Hui Yan, Rui Wang, Suryavathi Viswanadhapalli, Christian Cervantes, Funan He, Shuai Wu, Azad Khosh, Weiwei Luo, Jingwei Wang, Maria J. Fernandez, et al.

    B cells express many protein ligands, yet their regulatory functions are incompletely understood. We profiled ligand expression across murine B sublineage cells, including those activated by defined receptor signals, and assessed their regulatory capacities and specificities through in silico analysis of ligand-receptor interactions. Consequently, we identified a B cell subset that expressed cytokine interleukin-27 (IL-27) and chemokine CXCL10. Through the IL-27–IL-27 receptor interaction, these IL-27/CXCL10-producing B cells targeted CD40-activated B cells in vitro and, upon induction by immunization and viral infection, optimized antibody responses and antiviral immunity in vivo. Also present in breast cancer tumors and retained there through CXCL10-CXCR3 interaction–mediated self-targeting, these cells promoted B cell PD-L1 expression and immune evasion. Mechanistically, Il27 and Cxcl10 transcription was induced by synergizing Toll-like receptor (TLR) and CD40 signals and driven by coinduced transcription factor BATF3, which directly targeted these genes. By applying a discovery framework focusing on regulatory cells, our findings expand the recognized scope of B cell regulatory functions.


  • Evaluation of modified frailty index for predicting post‐operative outcomes after lower‐extremity free‐flap reconstruction

    by Mustafa T. A. Khan, Aashish Rajesh, Lisandro Montorfano, Melinda Lue, Brian Wong Won, Howard T. Wang, Rayaad C. Hosein

    Background: Risk stratification for complex procedures such as microsurgical reconstruction of the lower extremities is an important part of preoperative planning and counseling. The aim of this study was to determine the effectiveness of the modified five‐item frailty index (5‐mFI) score, a validated tool for assessing risk in surgical patients, in predicting postoperative complications after lower extremity (LE) free flap reconstruction.Methods: A retrospective review of the American College of Surgeons National Surgical Quality Improvement Program (ACS‐NSQIP) database was conducted from 2010 to 2020 on patients who underwent LE free‐flap reconstruction. 5‐mFI scores were calculated and patients were categorized as 5‐mFI ≥2 or p = .043) and hematologic complications (p = .033). In this population, there were also higher rates of reoperation (p = .003) and discharge to care facility (p < .001). Multivariable regression models further substantiated that 5‐mFI ≥2 was independently associated with increased overall complications [2.46, CI: 1.10–5.59, p = .031], hematologic complications [2.55, 1.02–6.35, p = .046], reoperation [4.55, 1.54–13.3, p = .006], and discharge to facility [2.86, 1.27–6.45, p = .011].Conclusions: There is a strong association of 5‐mFI ≥2 with adverse post‐operative outcomes in male patients undergoing LE free‐flap reconstruction. This can be a valuable adjunct in the counsel of patients for whom lower extremity salvage is feasible.


  • Intentional enterotomies: validation of a novel robotic surgery training exercise

    by Robert B. Laverty, Mustafa T. Khan, Ronit Patnaik, Christina S. Lee, Cassandra D. Leonardo, Robert W. Krell, Mamie C. Stull

    While laparoscopic simulation-based training is a well-established component of general surgery training, no such requirement or standardized curriculum exists for robotic surgery. Furthermore, there is a lack of high-fidelity electrocautery simulation training exercises in the literature. Using Messick’s validity framework, we sought to determine the content, response process, internal content and construct validity of a novel inanimate tissue model that utilizes electrocautery for potential incorporation in such curricula. A multi-institutional, prospective study involving medical students (MS) and general surgery residents (PGY1-3) was conducted. Participants performed an exercise using a biotissue bowel model on the da Vinci Xi robotic console during which they created an enterotomy using electrocautery, followed by approximation with interrupted sutures. Participant performance was recorded and then scored by crowd-sourced assessors of technical skill, along with three of the authors. Construct validity was determined via difference in Global Evaluative Assessment of Robotic Skills (GEARS) score, time to completion, and total number of errors between the two cohorts. Upon completion of the exercise, participants were surveyed on their perception of the exercise and its impact on their robotic training to determine content validity. 31 participants were enrolled and separated into two cohorts: MS + PGY1 vs. PGY2-3. Time spent on the robotic trainer (0.8 vs. 8.13 h, p = 0.002), number of bedside robotic assists (5.7 vs. 14.8, p < 0.001), and number of robotic cases as primary surgeon (0.3 vs. 13.1, p < 0.001) were statistically significant between the two groups. Differences in GEARS scores (18.5 vs. 19.9, p = 0.001), time to completion (26.1 vs. 14.4 min, p < 0.001), and total errors (21.5 vs. 11.9, p = 0.018) between the groups were statistically significant as well. Of the 23 participants that completed the post-exercise survey, 87% and 91.3% reported improvement in robotic surgical ability and confidence, respectively. On a 10-point Likert scale, respondents rated the realism of the exercise 7.5, educational benefit 9.1, and effectiveness in teaching robotic skills 8.7. Controlling for the upfront investment of certain training materials, each exercise iteration cost ~ $30. This study confirmed the content, response process, internal structure and construct validity of a novel, high-fidelity and cost-effective inanimate tissue exercise which successfully incorporates electrocautery. Consideration should be given to its addition to robotic surgery training programs.


  • The “Intern” Label: Introducing Unnecessary Confusion and Bias?

    by Ross E. Willis, Ronit Patnaik, Mustafa T. Khan, Claire Juhas, Anaya Shah

  • Gender Discrepancies in Middle Author Publications in US Academic General Surgery

    by Mustafa Tamim Alam Khan, Ronit Patnaik, Allison N. Laffoon, Lucijana Krokar, Spencer R. Ince, Estefania Hurtado, Mio Kitano, Danielle M. Fritze, Daniel L. Dent


  • Open Access

    Intentional Enterotomies: Validation of a Novel Robotic Surgery Training Exercise

    by Robert Burnham Laverty, Mustafa T. Khan, Ronit Patnaik, Christina Sooyeon Lee, Cassandra Dee Leonardo, Robert W. Krell, Mamie C. Stull

    Purpose: While laparoscopic simulation-based training is a well-established component of general surgery training, no such requirement or standardized curriculum exists for robotic surgery. Furthermore, there is a lack of high-fidelity electrocautery simulation training exercises in the literature. Using Messick’s validity framework, we sought to determine the content, response process, internal content and construct validity of a novel inanimate tissue model that utilizes electrocautery for potential incorporation in such curricula. Methods: Multi-institutional, prospective study involving medical students (MS) and general surgery residents (PGY1-3). Participants performed an exercise using a biotissue bowel model on the da Vinci Xi robotic console during which they created an enterotomy using electrocautery followed by approximation with interrupted sutures. Participant performance was recorded and then scored by crowd-sourced assessors of technical skill along with three of the authors. Construct validity was determined via difference in Global Evaluative Assessment of Robotic Skills (GEARS) score, time to completion and total number of errors between the two cohorts. Upon completion of the exercise, participants were surveyed on their perception of the exercise and its impact on their robotic training to determine content validity. Results: 31 participants were enrolled and separated into two cohorts: MS + PGY1 vs PGY2-3. Time spent on the robotic trainer (0.8 vs 8.13 hours, p=0.002), number of bedside robotic assists (5.7 vs 14.8, pConclusions: This study confirmed the content, response process, internal structure and construct validity of a novel, high-fidelity and cost-effective inanimate tissue exercise which successfully incorporates electrocautery. Consideration should be given to its addition to robotic surgery training programs.

  • Open Access

    Leukopenia is an independent risk factor for early postoperative complications following incision and drainage of anorectal abscess

    by Mustafa Tamim Alam Khan, Ronit Patnaik, Jian Yu Huang, Haisar Dao Campi, Lisandro Montorfano, Felice De Stefano, Raul J. Rosenthal, Steven D. Wexner

    Aim: Few data are available regarding the management of anorectal abscess in patients with leukopenia. The aim of this study was to investigate the impact of leukopenia among patients undergoing incision and drainage for anorectal abscess.Method: A retrospective review of the American College of Surgeons National Surgical Quality Improvement Program (ACS‐NSQIP) database identified patients from 2015 to 2020. Perianal fistulas and supralevator abscesses were excluded. Patients were grouped based on white blood cell (WBC) count: WBC < 4.5 cells/μl, WBC = 4.5–11.0 cells/μl and WBC > 11.0 cells/μl. The 30‐day overall complications and outcomes were compared using regression models, accounting for demographics and comorbidities.Results: Ten thousand two hundred and forty (70.3% male) patients were identified. Univariate analysis showed that, compared with patients with leukocytosis (WBC > 11.0 cells/μl) and normal WBC count (WBC = 4.5–11.0 cells/μl), patients with leukopenia (WBC p < 0.001), pulmonary (p < 0.001) and haematological complications (p < 0.001). They also had higher rates of readmission (p < 0.001), reoperation (p = 0.005), discharge to a care facility (p = 0.003), increased length of hospital stay (p = 0.004) and death (p < 0.001). Multivariable analysis identified leukopenia as an independent risk factor for overall complications [odds ratio (OR) 2.31, 95% CI 1.65–3.24; p < 0.001], pulmonary complications (OR 5.65, 95% CI 1.88–16.97; p = 0.002), haematological complications (OR 4.30, 95% CI 2.94–6.28; p < 0.001), unplanned readmission (OR 2.20, 95% CI 1.43–3.40; p < 0.001), reoperation (OR 1.80, 95% CI 1.10–2.93; p = 0.019) and death (OR 2.77, 95% CI 1.02–7.52; p = 0.046). Discharge to a care facility and length of stay were not significant on multivariable analysis.Conclusion: Leukopenia is associated with increased risk for pulmonary and haematological complications, readmissions, reoperations, discharge to a care facility and death after incision and drainage for anorectal abscess.