Comparative analysis of ChatGPT and Bard in answering pathology examination questions requiring image interpretation

OBJECTIVES

To evaluate the accuracy of ChatGPT and Bard in answering pathology examination questions requiring image interpretation.

METHODS

The study evaluated ChatGPT-4 and Bard's performance using 86 multiple-choice questions, with 17 (19.8%) focusing on general pathology and 69 (80.2%) on systemic pathology. Of these, 62 (72.1%) included microscopic images, and 57 (66.3%) were first-order questions focusing on diagnosing the disease. The authors presented these artificial intelligence (AI) tools with questions, both with and without clinical contexts, and assessed their answers against a reference standard set by pathologists.

RESULTS

ChatGPT-4 achieved a 100% (n = 86) accuracy rate in questions with clinical context, surpassing Bard's 87.2% (n = 75). Without context, the accuracy of both AI tools declined significantly, with ChatGPT-4 at 52.3% (n = 45) and Bard at 38.4% (n = 33). ChatGPT-4 consistently outperformed Bard across various categories, particularly in systemic pathology and first-order questions. A notable issue identified was Bard's tendency to "hallucinate" or provide plausible but incorrect answers, especially without clinical context.

CONCLUSIONS

This study demonstrated the potential of ChatGPT and Bard in pathology education, stressing the importance of clinical context for accurate AI interpretations of pathology images. It underlined the need for careful AI integration in medical education.

Histopathologic evaluation of gastric intestinal metaplasia in non-neoplastic biopsy specimens: Accuracy and interobserver reliability among general pathologists and pathology residents

OBJECTIVES

This study aimed to evaluate the accuracy and interobserver reliability of diagnosing and subtyping gastric intestinal metaplasia (IM) among general pathologists and pathology residents at a university hospital in Thailand, focusing on the challenges in the histopathologic evaluation of gastric IM for less experienced practitioners.

METHODS

The study analyzed 44 non-neoplastic gastric biopsies, using a consensus diagnosis of gastrointestinal pathologists as the reference standard. Participants included 6 general pathologists and 9 pathology residents who assessed gastric IM and categorized its subtype (complete, incomplete, or mixed) on digital slides. After initial evaluations and receiving feedback, participants reviewed specific images of gastric IM, as agreed by experts. Following a one-month washout period, a reevaluation of the slides was conducted.

RESULTS

Diagnostic accuracy, interobserver reliability, and time taken for diagnosis improved following training, with general pathologists showing higher accuracies than residents (median accuracy of gastric IM detection: 100 % vs. 97.7 %). Increased years of experience were associated with more IM detection accuracy (p-value<0.05). However, the overall median accuracy for diagnosing incomplete IM remained lower than for complete IM (86.4 % vs. 97.7 %). After training, diagnostic errors occurred in 6 out of 44 specimens (13.6 %), reported by over 40 % of participants. Errors involved omitting 5 slides with incomplete IM and 1 with complete IM, all showing a subtle presence of IM.

CONCLUSIONS

The study highlights the diagnostic challenges in identifying incomplete gastric IM, showing notable discrepancies in accuracy and interobserver agreement. It underscores the need for better diagnostic protocols and training to enhance detection and management outcomes.

Accuracy of GPT-4 in histopathological image detection and classification of colorectal adenomas

AIMS

To evaluate the accuracy of Chat Generative Pre-trained Transformer (ChatGPT) powered by GPT-4 in histopathological image detection and classification of colorectal adenomas using the diagnostic consensus provided by pathologists as a reference standard.

METHODS

A study was conducted with 100 colorectal polyp photomicrographs, comprising an equal number of adenomas and non-adenomas, classified by two pathologists. These images were analysed by classic GPT-4 for 1 time in October 2023 and custom GPT-4 for 20 times in December 2023. GPT-4's responses were compared against the reference standard through statistical measures to evaluate its proficiency in histopathological diagnosis, with the pathologists further assessing the model's descriptive accuracy.

RESULTS

GPT-4 demonstrated a median sensitivity of 74% and specificity of 36% for adenoma detection. The median accuracy of polyp classification varied, ranging from 16% for non-specific changes to 36% for tubular adenomas. Its diagnostic consistency, indicated by low kappa values ranging from 0.06 to 0.11, suggested only poor to slight agreement. All of the microscopic descriptions corresponded with their diagnoses. GPT-4 also commented about the limitations in its diagnoses (eg, slide diagnosis best done by pathologists, the inadequacy of single-image diagnostic conclusions, the need for clinical data and a higher magnification view).

CONCLUSIONS

GPT-4 showed high sensitivity but low specificity in detecting adenomas and varied accuracy for polyp classification. However, its diagnostic consistency was low. This artificial intelligence tool acknowledged its diagnostic limitations, emphasising the need for a pathologist's expertise and additional clinical context.

Learning at a distance: results of an international survey on the adoption of virtual conferences and whole slide imaging by pathologists

AIMS

This study presents the findings of a global survey of pathologists' views of online conferences and digital pathology.

METHODS

An online anonymous survey consisting of 11 questions focusing on pathologists' perceptions of virtual conferences and digital slides was distributed to practising pathologists and trainees across the globe using the authors' social media accounts and professional society connections. Participants were asked to rank their preference for various aspects of pathology meetings on a 5-point Likert scale.

RESULTS

There were 562 respondents from 79 countries. Several advantages of virtual meetings were recognised, including that they are less expensive to attend than in-person meetings (mean 4.4), more convenient to attend remotely (mean 4.3) and more efficient due to no loss of time for travel (mean 4.3). The lack of networking was reported as the main disadvantage of virtual conferences (mean 4.0). Most respondents (n=450, 80.1%) preferred hybrid or virtual meetings. About two-thirds (n=356, 63.3%) had no concern regarding the use of virtual slides for educational purposes and viewed them as an acceptable substitute for glass slides.

CONCLUSIONS

Online meetings and whole slide imaging are viewed as valuable tools in pathology education. Virtual conferences allow affordable registration fees and flexibility for participants. However, networking opportunities are limited, meaning in-person meetings cannot be entirely replaced by virtual conferences. Hybrid meetings may be a solution to maximise the benefits of both virtual and in-person meetings.

Validation of whole slide imaging for frozen section diagnosis of lymph node metastasis: A retrospective study from a tertiary care hospital in Thailand

BACKGROUND

The use of whole slide imaging (WSI) for frozen section (FS) diagnosis is helpful, particularly in the context of pathologist shortages. However, there is minimal data on such usage in resource-limited settings. This study aims to validate the use of WSI for FS diagnosis of lymph node metastasis using a low-cost virtual microscope scanner with consumer-grade laptops at a tertiary care hospital in Thailand.

METHODS

FS slides were retrieved for which the clinical query was to evaluate lymph node metastasis. They were digitized by a virtual microscope scanner (MoticEasyScan, Hong Kong) using up to 40× optical magnification. Three observers with different pathology experience levels diagnosed each slide, reviewing glass slides (GS) followed by digital slides (DS) after two weeks of a wash out period. WSI and GS diagnoses were compared. The time used for scanning and diagnosis of each slide was recorded.

RESULTS

295 FS slides were retrieved and digitized. The first-time successful scanning rate was 93.6 %. The mean scanning time was 2 min per slide. Both intraobserver agreement and interobserver agreement of WSI and GS diagnoses were high (Cohen's K; kappa value >0.84). The time used for DS diagnosis decreased as the observer's experience with WSI increased.

CONCLUSIONS

Despite varying pathological experiences, observers using WSI provided accurate FS diagnoses of lymph node metastasis. The time required for DS diagnoses decreased with additional observer's experience with WSI. Therefore, a WSI system containing low-cost scanners and consumer-grade laptops could be used for FS services in hospital laboratories lacking pathologists.

Cytomegalovirus colitis in an immunocompromised patient presenting with massive lower gastrointestinal bleeding

A 48-year-old man with HIV infection (CD4 count = 84 cells/μL) experienced hematemesis and hematochezia. Colonoscopy revealed massive bleeding in the colon, in which the source of the bleeding could not be identified. A total colectomy was performed. A large superficial ulcer at the rectosigmoid colon was observed. Histologically, abundant cytomegalovirus (CMV)-infected cells were noted. The pathological diagnosis was CMV colitis.

Combined hepatocellular-cholangiocarcinoma and its mimickers: Diagnostic pitfalls in surgical pathology

BACKGROUND

The diagnosis of combined hepatocellular-cholangiocarcinoma (cHCC-CCA) requires histomorphological detection of both hepatocellular carcinoma (HCC) and intrahepatic cholangiocarcinoma (iCCA). However, these primary liver cancers (PLCs) have a wide variety of microscopic appearances resulting in difficulties and uncertainties in cHCC-CCA's diagnosis. This study aims to perform a clinicopathologic analysis on the diagnosis of PLCs at a tertiary referral hospital in Thailand using traditional morphologic studies.

METHODS

A 5-year retrospective analysis of pathologically diagnosed PLCs was conducted. Pathological features and clinical characteristics of cHCC-CCA and other PLCs with the histopathologic resemblance to cHCC-CCA were studied. The pathological diagnosis was rendered based on histomorphological context rather than immunoreactivity. A literature review containing diagnostic pitfalls of cHCC-CCA was carried out.

RESULTS

PLCs from a total of 295 patients were retrieved, and cHCC-CCA accounted for 1.4% (n = 4) of the malignancies. Histomorphological evaluation is the most reliable diagnostic modality for cHCC-CCA. Extremely uncommon variants of iCCA (i.e., mucinous iCCA and adenosquamous iCCA) and iCCA arising with hepatocellular nodular lesions (i.e., iCCA with nodular regenerative hyperplasia (NRH), and iCCA in cirrhosis) could have a histomorphologic resemblance to that of cHCC-CCA.

CONCLUSIONS

Although there has been an exceedingly high incidence of iCCA in Thailand, such a commonness is not valid for cHCC-CCA in our series. Rare forms of iCCA could have a morphologic resemblance to that of cHCC-CCA. Regardless of the differentiation and immunophenotype, iCCA without a distinct HCC component should never be diagnosed as cHCC-CCA.

Pathological aspects of mucinous cholangiocarcinoma: A single-center experience and systematic review

Cholangiocarcinoma is a malignant neoplasm originating from the biliary epithelium. Its incidence is highest in Southeast Asia, especially in Thailand. Mucinous intrahepatic cholangiocarcinoma (mucinous iCCA), characterized by an abundant extracellular mucin pool accounting for at least 50% of total tumor volume, is an extremely rare variant of such malignancy and is notorious for rapid progression and dismal prognosis. We conducted an 11-year retrospective analysis of resected mucinous iCCAs from our institution with a systematic review on mucinous iCCAs and combined hepatocellular-mucinous cholangiocarcinoma (cHCC-mCCA). There were four resected mucinous iCCA specimens at our institution (prevalence = 0.5%). Most of the patients were male. The clinicopathological characteristics were variable. The diagnosis of mucinous iCCAs could not be rendered without pathological evaluation. Either intraductal papillary neoplasm or biliary intraepithelial neoplasia was present in three out of four cases. One patient passed away at 11 months following liver resection. A total of 19 mucinous iCCAs and four cHCC-mCCAs from previously published literature were analyzed. The 1-year mortality rate of mucinous iCCAs from our series and published literature is 35%. The present study confirmed that mucinous iCCA is an exceedingly uncommon variant of iCCA. The differential diagnoses include metastatic carcinoma with mucinous component and cHCC-mCCA.

The use of virtual pathology in teaching medical students: first experience of a medical school in Thailand

This article was migrated. The article was not marked as recommended. Background and objective:Conventional light microscopy (CLM) has long been used as a teaching modality in pathology courses. Virtual microscopy (VM) has recently challenged traditional practices and revolutionized pathology education. However, most medical schools in Thailand still use CLM to teach pathology to medical students. This study aims to evaluate the effectiveness of VM in terms of undergraduate teaching as the primary experience at Thammasat University.Methods:Participants were 29 second-year medical students who sat in the practical session of the alimentary system pathology using VM as teaching modality. Students took pre- and post-tests with 12 questions covering common alimentary system diseases, followed by 10 survey questions with a five-point Likert-style scale. Test and satisfaction scores were analyzed.Results:There was a significant difference (p<0.01) between student pre-test scores (mean ± SD, 1.7 ± 2.1) and post-test scores (7.1 ± 3.4). VM was viewed as a preferred learning modality, mainly because of its portability (mean: 4.9 on the five-point Likert-scale), satisfactory quality of images (4.4), permitting learning in less time (4.3), and stimulating cooperation between students while improving interaction with teachers (4).Conclusions:Student perceptions of VM as a new teaching-learning tool and their outstanding performance on the knowledge exam suggest that VM has a potential role in undergraduate teaching. The results of this study help to improve pathology teaching strategy in the integrated curriculum.

A case report of pulmonary amyloidosis recognized by detection of AA amyloid exclusively in alveolar macrophages

Amyloidosis is a rare condition in which tissue deposits of inert fibrillar protein result in organ damage and dysfunction. There are several types of amyloid fibrils. Some of the most common forms are AL (amyloid light chain) protein and AA (amyloid-associated) type of amyloid fibril protein. Pulmonary amyloidosis is relatively common but is usually asymptomatic. Thus, the diagnosis may be easily overlooked. A 78-year-old male with a history of multiple myeloma followed by systemic amyloidosis presented with abnormal chest CT showing diffuse interlobular thickening in the whole lung field with bilateral pleural effusion. Bronchoalveolar lavage and transbronchial biopsy were performed. Due to the patient's poor condition and hemorrhage, only one fragment was available from forceps biopsy. Histologically, there was no amyloid deposition in the lung parenchyma; however, some histiocytes showed eosinophilic granular contents which prompted us to perform additional staining. The cytoplasmic material turned to be positive with direct fast scarlet (DFS) staining and AA amyloid immunostaining. Similar macrophages with AA amyloid were also found in the bronchoalveolar fluid. We experienced a case with AA amyloidosis affecting the lung diagnosed by the presence of intracytoplasmic amyloid in alveolar macrophages. The microscopic changes were so subtle that they may be overlooked. Recognition of amyloid deposition in alveolar macrophages may be an important clue to diagnose pulmonary amyloidosis. Such finding is of particular significance in the small-sized specimens, such as biopsies and cytologic smears.