A Novel Approach to Ovarian Cancer Diagnosis via CT Imaging - GPT-4o-Driven Automated Feature Recognition and Validation in Clinical Settings

Sep 22, 2026

Upload by Mabel

GPT-4o is evaluated for non-invasive early-stage ovarian cancer diagnosis using computed tomography (CT) imaging when biomarker availability is limited and conventional AI approaches require specialized engineering knowledge. The model is trained to recognize ovarian lesions, extract key CT features, and output benign or malignant diagnoses. Radiologists and gynecologic oncologists independently review reports to assess performance across datasets, reliability in feature detection, and diagnostic gains in clinical settings.

LanguageEnglish

CategoryHealthcare

FormatPDF

Pages11

Reading Time28 min

AccessFree to download

Abstract

Background

Methods

Results

Conclusions

Study Context and Clinical Need

Importance of Early Detection

Diagnostic Gold Standard and Role of CT

[https://doi.org/10.1245/s10434-026-19248-2](https://doi.org/10.1245/s10434-026-19248-2) Annals of Surgical Oncology (2026) 33:6618–6628 ORIGINAL ARTICLE GYNECOLOGIC ONCOLOGY A Novel Approach to Ovarian Cancer Diagnosis via CT Imaging: GPT‑4o‑Driven Automated Feature Recognition and Validation in Clinical Settings Shimin Zhang, MD1 · Qiuyang Hou, MD2 · Mufei Ding, MS3 · Yuming Zhu, MD3 · Gang Dai, MD2 · Zhao Lu, MD4 · Zhuonan Liu, BS4 · Bosinan Chen, BM5 · Xiaogeng Li, BS6 · Jingyi Liu, BM5 · Kexue Deng, MD2 · Jiangdian Song, PhD3 · Xin Zhou, MD1 Received: 19 November 2025 / Accepted: 28 January 2026 / Published online: 17 February 2026 © The Author(s) 2026 Abstract Background Accurate non-invasive diagnosis of early-stage ovarian cancer remains challenging because of the limited number of biomarkers. Although artificial intelligence algorithms show promise for ovarian cancer diagnosis, their reliance on specialized engineering knowledge hinders their accessibility. The recent emergence of visual large language models such as GPT-4o has further expanded the potential of AI in this domain. Methods GPT-4o was trained to automatically recognize ovarian lesions, report key computed tomography (CT) features of ovarian lesions, and make a benign or malignant diagnosis based on these features. Radiologists and gynecologic oncologists independently reviewed the GPT-4o reports and evaluated GPT-4o's performance. Results GPT-4o achieved diagnostic accuracies of 80.80%, 79.14%, and 93.33% in the three datasets. Its performance surpassed that of gynecologic oncologist with 10 years of experience but was inferior to that of gynecologic oncologist with 16 years of experience and radiologists with ≥ 7 years of experience. The clinician-rated reliability in detecting the four key CT features was 4.22/5.00 for cyst wall and septum status; 4.24/5.00 for nodular or papillary protrusions; 4.30/5.00 for density and enhancement distribution; and 4.25/5.00 for cystic-solid characteristics. The use of GPT-4o increased the accuracy of radiologist and gynecologic oncologist diagnoses by 1.96% and 10.50%, respectively. Conclusions GPT-4o identifies the key CT features of ovarian cancer and achieves promising diagnostic accuracy with high-quality diagnostic evidence. Keywords Ovarian cancer · Artificial intelligence · Large language models · Diagnosis · Medical imaging Ovarian cancer is the eighth most common cancer in women and is a significant global health concern, accounting for 3.4% of new cancer cases and 4. 8% of cancer-related deaths in 2022.1 Ovarian cancer has the highest mortality rate among all gynecological cancers because patients often present at an advanced stage, as early symptoms are non-specific and reliable early detection biomarkers are Shimin Zhang, Qiuyang Hou, Mufei Ding, and Yuming Zhu have contributed equally to this work. * Kexue Deng, MD [dengkexue-anhui@163.com](dengkexue-anhui@163.com) * Jiangdian Song, PhD [song.jd0910@gmail.com](song.jd0910@gmail.com) * Xin Zhou, MD [xzhou@cmu.edu.cn](xzhou@cmu.edu.cn) 1 Department of Obstetrics and Gynecology, Shengjing Hospital of China Medical University, Shenyang, Liaoning, China 2 Department of Radiology, The First Affiliated Hospital of University of Science and Technology of China (USTC); Division of Life Sciences and Medicine, USTC, Hefei, Anhui, China 3 School of Health Management, China Medical University, Shenyang, Liaoning, China 4 Department of Radiology, Shengjing Hospital of China Medical University, Shenyang, Liaoning, China 5 The First Hospital of China Medical University, Shenyang, Liaoning, China 6 School of Intelligent Medicine, China Medical University, Shenyang, Liaoning, China A Radiologists B M B Key features detection and malignancy estimation GUIDE GPT-4o C Radiologists Gynecologic oncologists | | R1 | R2 | R3 | G1 | ... | | --- | --- | --- | --- | --- | --- | | Feature 1 | 4 | 3 | ... | ... | .

FAQs

What problem does GPT-4o aim to address in ovarian cancer diagnosis?

Early-stage ovarian cancer diagnosis is difficult due to limited biomarkers, and many AI methods are hard to access because they depend on specialized engineering knowledge. GPT-4o targets automated, non-invasive assessment using CT imaging features.

How was the GPT-4o diagnostic process evaluated in the study?

GPT-4o was trained to recognize ovarian lesions and report key CT features, then provide benign or malignant diagnoses. Radiologists and gynecologic oncologists independently reviewed the GPT-4o reports and assessed performance.

What improvements and reliability results were reported for GPT-4o?

GPT-4o achieved diagnostic accuracies across three datasets and demonstrated reliability scores for detecting four key CT features. The study also reports that GPT-4o increased radiologist and gynecologic oncologist diagnostic accuracy compared with clinician-only assessments.

Sep 22, 2026

Upload by Mabel