Minahil Cheema

Ms. Minahil Cheema

orcid

Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
9 October 2025 in PLOS Digital Health
Verified
Journal Article
Open access
PDF file
YSYahya Shaikh
Yahya Shaikh
ZJZainab Asiyah Jeelani-Shaikh
Zainab Asiyah Jeelani-Shaikh
MJMuzamillah Mushtaq Jeelani
Muzamillah Mushtaq Jeelani
AJAamir Javaid
Aamir Javaid
TMTauhid Mahmud
Tauhid Mahmud
ZSZishan Siddiqui
Zishan Siddiqui
The stochastic nature of next-token generation and resulting response variability in Large Language Models (LLMs) outputs pose challenges in ensuring consistency and accuracy on knowledge assessments. This study introduces a novel multi-agent framework, referred to as a “Council of AIs”, to enhance LLM performance through collaborative decision-making. The Council consists of multiple GPT-4 instances that iteratively discuss and reach consensus on answers facilitated by a designated “Facilitator AI.” This methodology was applied to 325 United States Medical Licensing Exam (USMLE) questions across all three exam stages: Step 1, focusing on biomedical sciences; Step 2 evaluating clinical knowledge (CK)\; and Step 3, evaluating readiness for independent medical practice. The Council achieved consensus that were correct 97%, 93%, and 94% of the time for Step 1, Step 2 CK, and Step 3, respectively, outperforming single-instance GPT-4 models. In cases where there wasn’t an initial unanimous response, the Council deliberations achieved a consensus that was the correct answer 83% of the time, with the Council correcting over half (53%) of the responses that majority vote had gotten incorrect. The odds of a majority voting response changing from incorrect to correct were 5 (95% CI: 1.1, 22.8) times higher than the odds of changing from correct to incorrect after discussion. This study provides the first evidence that the semantic entropy of the response space can consistently be reduced to zero—demonstrated here through Council deliberation, and suggesting the possibility of other mechanisms to achieve the same outcome.. This study revealed that in a Council model, response variability, often considered a limitation, can be transformed into a strength that supports adaptive reasoning and collaborative refinement of answers. These findings suggest new paradigms for AI implementation and reveal the heightened strength that emerges when AIs begin to collaborate as a collective rather than operate alone.

5 March 2024 in Microbiology Spectrum
Verified
Journal Article
Open access
MVMary Jane E. Vaeth
Mary Jane E. Vaeth
MCMinahil Cheema
Minahil Cheema
SOSarah Omer
Sarah Omer
IGIshaan Gupta
Ishaan Gupta
KSKristie J. Sun
Kristie J. Sun
ZSZishan K. Siddiqui
Zishan K. Siddiqui
We conducted a single-center study at a free community testing site in Baltimore City to assess the accuracy of self-performed rapid antigen tests (RATs) for COVID-19. Self-administered BinaxNOW RATs were compared with clinician-performed RATs and against a reference lab molecular testing as the gold standard. Of the 953 participants, 14.9% were positive for SARS- CoV-2 as determined by RT-PCR. The sensitivity and specificity were similar for both self- and clinician-performed RATs (sensitivity: 83.9% vs 88.2%, P = 0.40; specificity: 99.8% vs 99.6%, P = 0.6). Subgroup comparisons based on age and race yielded similar results. Notably, 5.2% (95% CI: 1.5% to 9.5%) of positive results were potentially missed due to participant misinterpretation of the self-test card. However, the false-positive rate for RATs was reassuringly comparable in accuracy to clinician-administered tests. These findings hold significant implications for physicians prescribing treatment based on patient-reported, self-administered positive test results. Our study provides robust evidence supporting the reliability and utility of patient-performed RATs, underscoring their comparable accuracy to clinician-performed RATs, and endorsing their continued use in managing COVID-19. Further studies using other rapid antigen test brands are warranted. IMPORTANCE: Accurate and accessible COVID-19 testing is crucial for effective disease control and management. A recent single-center study conducted in Baltimore City examined the reliability of self-performed rapid antigen tests (RATs) for COVID-19. The study found that self-administered RATs yielded similar sensitivity and specificity to clinician-performed tests, demonstrating their comparable accuracy. These findings hold significant implications for physicians relying on patient-reported positive test results for treatment decisions. The study provides robust evidence supporting the reliability and utility of patient-performed RATs, endorsing their continued use in managing COVID-19. Furthermore, the study highlights the need for further research using different rapid antigen test brands to enhance generalizability. Ensuring affordable and widespread access to self-tests is crucial, particularly in preparation for future respiratory virus seasons and potential waves of reinfection of SARS-CoV-2 variants such as the Omicron variant.

9 March 2023 in medRxiv
Verified
Preprint
Open access
PDF file
MVMary Jane E. Vaeth
Mary Jane E. Vaeth
AMAsia Mitchell
Asia Mitchell
KSKristie Sun
Kristie Sun
MFMaisha Foyez
Maisha Foyez
LBLourdes Byrd
Lourdes Byrd
ZSZishan Siddiqui
Zishan Siddiqui
Rapid antigen tests are widely used to diagnose infection with SARS-CoV-2, and millions of kits have been distributed for free by government agencies. However, unused and expired kits beyond their final expiration dates remain prevalent in people’s homes. This study aimed to determine the accuracy of expired BinaxNOW COVID rapid antigen test kits. 100 expired and 100 unexpired test kits were checked for sensitivity and specificity using positive and negative controls, respectively. The results showed that there was no change in the sensitivity and specificity of BinaxNOW COVID rapid antigen test kits four months beyond the manufacturer-extended expiration date when using manufacturer-provided positive controls. The findings provide confidence in the accuracy of expired test kits, which could potentially reduce waste and strengthen supply chain resilience for pandemic preparedness. Further research utilizing actual human specimens can help determine the true accuracy of expired rapid antigen test kits in clinical use.

17 August 2023 in Microbiology Spectrum
Verified
Journal Article
Open access
MVMary Jane E. Vaeth
Mary Jane E. Vaeth
OAOmar Abdullah
Omar Abdullah
MCMinahil Cheema
Minahil Cheema
KSKristie Sun
Kristie Sun
MEMaryam Elhabashy
Maryam Elhabashy
ZSZishan K. Siddiqui
Zishan K. Siddiqui
The study examining the reliability of expired antigen testing kits in the context of COVID-19 has significant real-world implications. By demonstrating that these expired kits retain their sensitivity in detecting the virus, this work provides evidence that expired kits can still be utilized, reducing waste and optimizing resources in health care systems.

30 January 2025 in Critical Care Research and Practice
Verified
Journal Article
Open access
CLCassidy Lavin
Cassidy Lavin
JEJacob Epstein
Jacob Epstein
HCHuanwen Chen
Huanwen Chen
MCMinahil Cheema
Minahil Cheema
JYJerry Yang
Jerry Yang
QTQuincy Tran
Quincy Tran
Objective: To investigate differences in outcomes among patients with acute ischemic stroke from large vessel occlusion (AIS‐LVO) transferred from rural and urban hospitals to University of Maryland Medical Center (UMMC) for mechanical thrombectomy (MT). Methods: We identified patients with AIS‐LVO transferred to UMMC for MT from July 2016 to June 2023. Primary outcome was good neurologic outcome, defined as 90‐day modified Rankin score 0–2. Multivariable logistic regression was used to identify predictors for the primary outcome. Results: We analyzed 526 patients, 233 (44%) transferred from rural hospitals in Maryland. Median NIHSS was 17 [IQR 14–20] and was similar between groups. Patients from state‐designated rural hospitals were transferred from a longer distance (difference of 57.8 km, p = 0.001), but had shorter intervals from last known well time to recanalization (difference 19 min, p = 0.24). They had similar odds of good neurologic outcome (OR 0.88, 95% CI 0.43–1.78, p = 0.72). Conclusions: Patients transferred from rural areas for MT for AIS‐LVO, at our institution, had a similar likelihood of achieving 90‐day mRS 0–2 as those transferred from urban areas.

25 September 2024 in Western Journal of Emergency Medicine
Verified
Journal Article
Open access
PDF file
PMPatrick McGinnis
Patrick McGinnis
SFSamantha Camp Ferrell
Samantha Camp Ferrell
MCMinahil Cheema
Minahil Cheema
SJShriya Jaddu
Shriya Jaddu
QTQuincy Tran
Quincy Tran
JDJessica Downing
Jessica Downing
Author(s): McGinnis, Patrick; Camp, Samantha; Cheema, Minahil; Jaddu, Shriya; Tran, Quincy K; Downing, Jessica