Automated detection of cancer cells in effusion specimens by DNA karyometry

A histopathological image classification method for cholangiocarcinoma based on spatial-channel feature fusion convolution neural network

Histopathological image analysis plays an important role in the diagnosis and treatment of cholangiocarcinoma. This time-consuming and complex process is currently performed manually by pathologists. To reduce the burden on pathologists, this paper proposes a histopathological image classification method for cholangiocarcinoma based on spatial-channel feature fusion convolutional neural networks. Specifically, the proposed model consists of a spatial branch and a channel branch. In the spatial branch, residual structural blocks are used to extract deep spatial features. In the channel branch, a multi-scale feature extraction module and some multi-level feature extraction modules are designed to extract channel features in order to increase the representational ability of the model. The experimental results of the Multidimensional Choledoch Database show that the proposed method performs better than other classical CNN classification methods.

DNA image cytometry ploidy analysis technique improves the detection rate of pleural effusion cytology

OBJECTIVE

To explore the clinical diagnostic value of DNA image cytometry (DNA-ICM) ploidy analysis in malignant pleural effusion cancer screening, this study analyzed the effect of exfoliated cell smears (ECSs), cell blocks (CBs), and immunochemistry.

METHOD

A total of 830 cases of pleural effusion were considered for the DNA-ICM ploidy analysis. The ECSs were centrifuged, the CBs were formed, and the DNA-ICM ploidy analysis was carried out in the diagnosis of malignant pleural effusion. Immunochemistry and biopsy was applied to differentiate between benign and malignant pleural effusion and to determine the source of the latter. The sensitivity and specificity differences between the three methods alone and in combination were compared.

RESULTS

The sensitivity of the DNA-ICM, ECS, and CB methods was 96.28%, 94.93%, and 95.95%, respectively, and the specificity of each method was 86.52%, 87.08%, and 86.14%, respectively. The sensitivity and specificity of the combined diagnosis method were 99.32% and 75.09%, respectively. Among the 22 cases diagnosed as positive in the DNA-ICM ploidy analysis but negative in the ECS and CB analyses, four cases were diagnosed as positive by comprehensive clinical diagnosis.

CONCLUSION

The sensitivity and specificity of DNA-ICM ploidy analysis are high; the positive detection rate of pleural fluid cytology is effectively increased, and the missed detection rate of cell pathologies is effectively reduced. The combination of the three methods significantly improves the specificity and sensitivity of the diagnosis of malignant pleural effusion, and immunochemistry with CBs can be used to accurately analyze the primary tumor site.

DNA Karyometry for Automated Detection of Cancer Cells

Background: Microscopical screening of cytological samples for the presence of cancer cells at high throughput with sufficient diagnostic accuracy requires highly specialized personnel which is not available in most countries. Methods: Using commercially available automated microscope-based screeners (MotiCyte and EasyScan), software was developed which is able to classify Feulgen-stained nuclei into eight diagnostically relevant types, using supervised machine learning. the nuclei belonging to normal cells were used for internal calibration of the nuclear DNA content while nuclei belonging to those suspicious of being malignant were specifically identified. The percentage of morphologically abnormal nuclei was used to identify samples suspected of malignancy, and the proof of DNA-aneuploidy was used to definitely determine the state malignancy. A blinded study was performed using oral smears from 92 patients with Fanconi anemia, revealing oral leukoplakias or erythroplakias. In an earlier study, we compared diagnostic accuracies on 121 serous effusion specimens. In addition, using a blinded study employing 80 patients with prostate cancer who were under active surveillance, we aimed to identify those whose cancers would not advance within 4 years. Results: Applying a threshold of the presence of >4% of morphologically abnormal nuclei from oral squamous cells and DNA single-cell or stemline aneuploidy to identify samples suspected of malignancy, an overall diagnostic accuracy of 91.3% was found as compared with 75.0% accuracy determined by conventional subjective cytological assessment using the same slides. Accuracy of automated screening effusions was 84.3% as compared to 95.9% of conventional cytology. No prostate cancer patients under active surveillance, revealing DNA-grade 1, showed progress of their disease within 4.1 years. Conclusions: An automated microscope-based screener was developed which is able to identify malignant cells in different types of human specimens with a diagnostic accuracy comparable with subjective cytological assessment. Early prostate cancers which do not progress despite applying any therapy could be identified using this automated approach.

Artificial intelligence assisted cytological detection for early esophageal squamous epithelial lesions by using low-grade squamous intraepithelial lesion as diagnostic threshold

BACKGROUND

Manual cytological diagnosis for early esophageal squamous cell carcinoma (early ESCC) and high-grade intraepithelial neoplasia (HGIN) is unsatisfactory. Herein, we have introduced an artificial intelligence (AI)-assisted cytological diagnosis for such lesions.

METHODS

Low-grade squamous intraepithelial lesion or worse was set as the diagnostic threshold for AI-assisted diagnosis. The performance of AI-assisted diagnosis was evaluated and compared to that of manual diagnosis. Feasibility in large-scale screening was also assessed.

RESULTS

AI-assisted diagnosis for abnormal cells was superior to manual reading by presenting a higher efficiency for each slide (50.9 ± 0.8 s vs 236.8 ± 3.9 s, p = 1.52 × 10^-76 ) and a better interobserver agreement (93.27% [95% CI, 92.76%-93.74%] vs 65.29% [95% CI, 64.35%-66.22%], p = 1.03 × 10^-84 ). AI-assisted detection showed a higher diagnostic accuracy (96.89% [92.38%-98.57%] vs 72.54% [65.85%-78.35%], p = 1.42 × 10^-14 ), sensitivity (99.35% [95.92%-99.97%] vs 68.39% [60.36%-75.48%], p = 7.11 × 10^-15 ), and negative predictive value (NPV) (97.06% [82.95%-99.85%] vs 40.96% [30.46%-52.31%], p = 1.42 × 10^-14 ). Specificity and positive predictive value (PPV) were not significantly differed. AI-assisted diagnosis demonstrated a smaller proportion of participants of interest (3.73%, [79/2117] vs.12.84% [272/2117], p = 1.59 × 10^-58 ), a higher consistence between cytology and endoscopy (40.51% [32/79] vs. 12.13% [33/272], p = 1.54 × 10^- 8), specificity (97.74% [96.98%-98.32%] vs 88.52% [87.05%-89.84%], p = 3.19 × 10^-58 ), and PPV (40.51% [29.79%-52.15%] vs 12.13% [8.61%-16.75%], p = 1.54 × 10^-8 ) in community-based screening. Sensitivity and NPV were not significantly differed. AI-assisted diagnosis as primary screening significantly reduced average cost for detecting positive cases.

CONCLUSION

Our study provides a novel cytological method for detecting and screening early ESCC and HGIN.

DNA image cytometric analysis of bronchial washings as an adjunct for the detection of lung cancer in a clinical setting

Background

DNA aneuploidy has a potential to become an adjunct to conventional cytology for diagnosis of lung cancer, but its value in bronchial washings has not been well evaluated.

Methods

We conducted a retrospective study on patients who underwent bronchoscopy and the bronchial washings were submitted for both cytology and DNA image cytometry (DNA‐ICM) examination. The sensitivity and specificity of two methods and both in combination were compared. Analysis of clinical subgroups and DNA histogram were also performed.

Results

The study included 626 patients (326 patients with lung cancer and 300 patients with benign lung diseases). The sensitivity of cytology, DNA‐ICM, and combination test for lung cancer were 53.3%, 62.3%, and 75.8%, respectively, and the sensitivity of DNA‐ICM and combination test were superior to that of cytology (p < 0.05). A modest reduction of specificity was found in DNA‐ICM compared with cytology (91.3% vs. 98.3%, p < 0.05). Subgroup analysis showed there was no significant difference in sensitivity of DNA‐ICM between the visible tumor group and the invisible tumor group (66.5% vs. 56.9%, χ² = 3.114, p = 0.078). Among 101 patients with invisible endobronchial tumor, the positive rates for DNA‐ICM of washing, cytology of washing, brushing and biopsy were 62.4%, 41.6%, 40.6%, and 45.5%, respectively. DNA‐ICM in combination with the basic bronchoscopy techniques could increase the sensitivity from 67.3% to 87.1% (p = 0.000). The DNA histogram analysis showed 25.3% washing samples of lung cancer were diploid pattern, 49.4% were scattered aneuploid cells pattern, and 25.3% were aneuploid peaks pattern. Small cell lung cancer had the highest proportion of aneuploid peaks pattern (p < 0.05).

Conclusions

DNA‐ICM could be used as an adjunct for the detection of lung cancer. The combination of DNA‐ICM and basic bronchoscopy techniques could significantly increase the sensitivity, especially for the patients suspected of peripheral lung cancer, and contribute to select subjects for advanced bronchoscopy.

A Rapid Cytological Screening as pre-Endoscopy Screening for Early Esophageal Squamous Cell Lesions: A Prospective Pilot Study from a Chinese Academic Center

Background: Cytological detection of early esophageal squamous cell carcinoma (ESCC) remains challenging. Therefore, we introduced a rapid cytological screening method and evaluated its efficacy as a pre-endoscopy screening for early ESCC and precursor lesions. Methods: This method consisted of a sponge sample retrieval, automatic liquid-based cytological treatment and slides preparation, computer-assisted screening and manual diagnosis. Efficacy for detection of early ESCC and precursor lesions was evaluated. Also, diagnostic efficiency was compared with manual diagnosis. Results: Eighty-three patients with early ESCC and precursor lesions and 2,090 asymptomatic participants with high risks of ESCC were enrolled. Whole procedure was accomplished within two working days. Abnormal cells were detected in all 83 patients, and in 272 (13.01%) subjects among 2,090 asymptomatic participants. Early ESCC, high-grade intraepithelial neoplasia, low-grade intraepithelial neoplasia and reflux esophagitis and normal endoscopic findings were detected in 8, 13, 11, 187 and 53 participants with abnormal cells, respectively. The calculated sensitivity, specificity, positive predictive value and negative predictive value for detection of early ESCC and precursor lesions were 100%, 88.34%, 11.76%, and 100%, respectively. Compared with manual diagnosis, this method was accomplished in a shorter time duration (5.4 ± 0.45 min vs 320.2 ± 132.4 min, p < 0.001), a higher diagnostic accuracy (96.7% vs74.4%, p = 0.015) and a better inter-observer agreement (93.3% vs66.7%, K = 0.286, p < 0.001). Conclusions: Our study provides a promising methodology as pre-endoscopy screening for early ESCC and precursor lesions.

What is the state of the art of computer vision-assisted cytology? A Systematic Literature Review

Cytology is a low-cost and non-invasive diagnostic procedure employed to support the diagnosis of a broad range of pathologies. Cells are harvested from tissues by aspiration or scraping, and it is still predominantly performed manually by medical or laboratory professionals extensively trained for this purpose. It is a time-consuming and repetitive process where many diagnostic criteria are subjective and vulnerable to human interpretation. Computer Vision technologies, by automatically generating quantitative and objective descriptions of examinations' contents, can help minimize the chances of misdiagnoses and shorten the time required for analysis. To identify the state-of-art of computer vision techniques currently applied to cytology, we conducted a Systematic Literature Review, searching for approaches for the segmentation, detection, quantification, and classification of cells and organelles using computer vision on cytology slides. We analyzed papers published in the last 4 years. The initial search was executed in September 2020 and resulted in 431 articles. After applying the inclusion/exclusion criteria, 157 papers remained, which we analyzed to build a picture of the tendencies and problems present in this research area, highlighting the computer vision methods, staining techniques, evaluation metrics, and the availability of the used datasets and computer code. As a result, we identified that the most used methods in the analyzed works are deep learning-based (70 papers), while fewer works employ classic computer vision only (101 papers). The most recurrent metric used for classification and object detection was the accuracy (33 papers and 5 papers), while for segmentation it was the Dice Similarity Coefficient (38 papers). Regarding staining techniques, Papanicolaou was the most employed one (130 papers), followed by H&E (20 papers) and Feulgen (5 papers). Twelve of the datasets used in the papers are publicly available, with the DTU/Herlev dataset being the most used one. We conclude that there still is a lack of high-quality datasets for many types of stains and most of the works are not mature enough to be applied in a daily clinical diagnostic routine. We also identified a growing tendency towards adopting deep learning-based approaches as the methods of choice.

The role of DNA image cytometry in screening oral potentially malignant lesions using brushings: A systematic review

It is believed that the majority of oral cancers develop from oral potentially malignant lesions (OPML). Though they can be easily detected during screening, risk stratification is difficult. During screening clinicians often find it difficult to distinguish OPMLs from benign lesions, and predicting OPML at risk of malignant transformation is particularly challenging. DNA aneuploidy has been known to be a marker of malignancy in a number of sites including the oral cavity. We performed a systematic review to evaluate the effectiveness of DNA-ICM using brushings in differentiating OPMLs from benign/inflammatory lesions during screening and in predicting malignant transformation. MEDLINE, Pubmed, EMBASE electronic databases were systematically searched using a combination of keywords and subject headings. A total of 11 articles satisfied our inclusion criteria. These studies reported a wide range of sensitivity (16-96.4%) and specificity (90-100%) due to the differences in study design, definitions of high risk or low risk lesions and DNA-ICM protocol used. No long-term longitudinal studies were identified to assess the role of DNA-ICM using brushings in predicting malignant transformation. No studies evaluated the role of DNA-ICM in community screening settings. A number of studies combined DNA-ICM with other techniques like cytology or argyrophilic nucleolar organizer region counts leading to improved test results. In spite of DNA aneuploidy being accepted as a marker of malignancy, there is limited evidence of DNA-ICM using brushings being successful as an adjunct oral cancer screening tool. Longitudinal studies and large community screening studies need to be undertaken to draw stronger conclusion.