Abstract:Liquid biopsy of cancers, detecting tumor-related information from liquid samples, has attracted wide attentions as an emerging technology. Our previously reported large-area PERFECT (Precise-Efficient-Robust-Flexible-Easy-Controllable-Thin) filter has demonstrated competitive sensitivity in recovering rare tumor cells from clinical samples. However, it is time-consuming and easily biased to manually inspect rare target cells among numerous background cells distributed in a large area (Φ ≥ 13 mm). This puts forward an urgent demand for rapid and bias-free inspection. Hereby, this paper implemented deep learning-based object detection for the inspection of rare tumor cells from large-field images of PERFECT filters with hematoxylin-eosin (HE)-stained cells recovered from bronchoalveolar lavage fluid (BALF). CenterNet, EfficientDet, and YOLOv5 were trained and validated with 240 and 60 image blocks containing tumor and/or background cells, respectively. YOLOv5 was selected as the basic network given the highest mAP@0.5 of 92.1%, compared to those of CenterNet and EfficientDet at 85.2% and 91.6%, respectively. Then, tricks including CIoU loss, image flip, mosaic, HSV augmentation and TTA were applied to enhance the performance of the YOLOv5 network, improving mAP@0.5 to 96.2%. This enhanced YOLOv5 network-based object detection, named as BALFilter Reader, was tested and cross-validated on 24 clinical cases. The overall diagnosis performance (~2 min) with sensitivity@66.7% ± 16.7%, specificity@100.0% ± 0.0% and accuracy@75.0% ± 12.5% was superior to that from two experienced pathologists (10-30 min) with sensitivity@61.1%, specificity@16.7% and accuracy@50.0%, with the histopathological result as the gold standard. The AUC of the BALFilter Reader is 0.84 ± 0.08. Moreover, a customized Web was developed for a user-friendly interface and the promotion of wide applications. The current results revealed that the developed BALFilter Reader is a rapid, bias-free and easily accessible AI-enabled tool to promote the transplantation of the BALFilter technique. This work can easily expand to other cytopathological diagnoses and improve the application value of micro/nanotechnology-based liquid biopsy in the era of intelligent pathology.

Learning an Improved Object Detection Approach Based on the YOLO Algorithm to Identify Circulating Tumor Cells

TSO-DETR: A Network for Small Object Detection of Cervical Cells in TCT Smear

Intelligent Tumor Cell Detection Method Based on Circulating Tumor Cell (CTC) Technology

A New Method for CTC Images Recognition Based on Machine Learning

Interpretatively Automated Identification of Circulating Tumor Cells from Human Peripheral Blood with High Performance

An Innovative Algorithm Combining Attention Mechanism and Feature Fusion for Circulating Tumor Cells Detection

A Model for the Detection of Circulating Tumor Cells Based on Micro-OCT Imaging

Improving Circulating Tumor Cell Detection Using Image Synthesis and Transformer Models in Cancer Diagnostics

Circulating Tumor Cell Identification Based on Deep Learning

Advancements in AI for Oncology: Developing an Enhanced YOLOv5-based Cancer Cell Detection System

Tumor Cell Small Object Detection Algorithm Based on Improved YOLOv5

Precisely Enumerating Circulating Tumor Cells Utilizing a Multi-Functional Microfluidic Chip and Unique Image Interpretation Algorithm.

Detection of cervical cancer cells in complex situation based on improved YOLOv3 network

Ellipsoidal Mirror Dark-Field Scanning for Detection of Circulating Tumor Cells

Enhanced YOLOv5 network-based object detection (BALFilter Reader) promotes PERFECT filter-enabled liquid biopsy of lung cancer from bronchoalveolar lavage fluid (BALF)

Improvement of Specific Detection of Circulating Tumor Cells Using Combined CD45 Staining and Fluorescence in Situ Hybridization

Homogeneous two-dimensional visual and fluorescence analysis of circulating tumor cells in clinical samples via steric hindrance regulated enzymes recognition cleavage and elongation

Machine Learning Assisted Dual-Modal SERS Detection for Circulating Tumor Cells

Research on Cervical Cancer p16/Ki-67 Immunohistochemical Dual-Staining Image Recognition Algorithm Based on YOLO

A novel dual-function SERS-based identification strategy for preliminary screening and accurate diagnosis of circulating tumor cells

A Novel Automatic Platform Based on Nanostructured Microfluidic Chip for Isolating and Identification of Circulating Tumor Cells