Colorectal polyps are important early markers of colorectal cancer, and their early detection is crucial for cancer prevention. Although existing polyp segmentation models have achieved certain results, they still face challenges such as diverse polyp morphology, blurred boundaries, and insufficient feature extraction. To address these issues, this study proposes a parallel coordinate fusion network (PCFNet), aiming to improve the accuracy and robustness of polyp segmentation. PCFNet integrates parallel convolutional modules and a coordinate attention mechanism, enabling the preservation of global feature information while precisely capturing detailed features, thereby effectively segmenting polyps with complex boundaries. Experimental results on Kvasir-SEG and CVC-ClinicDB demonstrate the outstanding performance of PCFNet across multiple metrics. Specifically, on the Kvasir-SEG dataset, PCFNet achieved an F1-score of 0.897 4 and a mean intersection over union (mIoU) of 0.835 8; on the CVC-ClinicDB dataset, it attained an F1-score of 0.939 8 and an mIoU of 0.892 3. Compared with other methods, PCFNet shows significant improvements across all performance metrics, particularly in multi-scale feature fusion and spatial information capture, demonstrating its innovativeness. The proposed method provides a more reliable AI-assisted diagnostic tool for early colorectal cancer screening.
Humans ; Colonic Polyps/diagnostic imaging* ; Colorectal Neoplasms/diagnostic imaging* ; Neural Networks, Computer ; Image Processing, Computer-Assisted/methods* ; Algorithms ; Early Detection of Cancer

To address the challenges in blood cell recognition caused by diverse morphology, dense distribution, and the abundance of small target information, this paper proposes a blood cell detection algorithm - the "You Only Look Once" model based on hybrid mixing attention and deep over-parameters (HADO-YOLO). First, a hybrid attention mechanism is introduced into the backbone network to enhance the model's sensitivity to detailed features. Second, the standard convolution layers with downsampling in the neck network are replaced with deep over-parameterized convolutions to expand the receptive field and improve feature representation. Finally, the detection head is decoupled to enhance the model's robustness for detecting abnormal cells. Experimental results on the Blood Cell Counting Dataset (BCCD) demonstrate that the HADO-YOLO algorithm achieves a mean average precision of 90.2% and a precision of 93.8%, outperforming the baseline YOLO model. Compared with existing blood cell detection methods, the proposed algorithm achieves state-of-the-art detection performance. In conclusion, HADO-YOLO offers a more efficient and accurate solution for identifying various types of blood cells, providing valuable technical support for future clinical diagnostic applications.
Algorithms ; Humans ; Blood Cells/cytology* ; Blood Cell Count/methods* ; Neural Networks, Computer ; Deep Learning ; Detection Algorithms

In order to overcome the shortcomings of high false positive rate and poor generalization in the detection of microcalcification clusters regions, this paper proposes a method combining discriminative deep belief networks (DDBNs) to automatically and quickly locate the regions of microcalcification clusters in mammograms. Firstly, the breast region was extracted and enhanced, and the enhanced breast region was segmented to overlapped sub-blocks. Then the sub-block was subjected to wavelet filtering. After that, DDBNs model for breast sub-block feature extraction and classification was constructed, and the pre-trained DDBNs was converted to deep neural networks (DNN) using a softmax classifier, and the network is fine-tuned by back propagation. Finally, the undetected mammogram was inputted to complete the location of suspicious lesions. By experimentally verifying 105 mammograms with microcalcifications from the Digital Database for Screening Mammography (DDSM), the method obtained a true positive rate of 99.45% and a false positive rate of 1.89%, and it only took about 16 s to detect a 2 888 × 4 680 image. The experimental results showed that the algorithm of this paper effectively reduced the false positive rate while ensuring a high positive rate. The detection of calcification clusters was highly consistent with expert marks, which provides a new research idea for the automatic detection of microcalcification clusters area in mammograms.
Algorithms ; Breast Neoplasms/diagnostic imaging* ; Calcinosis/diagnostic imaging* ; Early Detection of Cancer ; Humans ; Mammography ; Neural Networks, Computer

Objective To investigate the application value of eye tracking in lie detection. Methods The 40 subjects were randomly divided into two groups. The pupil diameter, fixation duration, points of fixation and blink frequency of the subjects in the experimental group in observing target stimulation and non-target stimulation were recorded with eye tracker after they accomplished the mock crime. The eye movement parameters of subjects in the control group were directly collected. The differences in eye movement parameters of the experimental group and the control group in observing target stimulation and non-target stimulation were analyzed by t-test. Pearson coefficient analysis of correlation between eye movement parameters that had differences was conducted. The effectiveness of eye movement parameters to distinguish between the experimental group and the control group was calculated by the receiver operator characteristic （ROC） curve. Results Participants from the experimental group had shorter average pupil diameter, longer average fixation duration and fewer fixation points （P<0.05）, but the differences in blink frequency had no statistical significance. The differences in the above indicators of the control group in observing target stimulation and non-target stimulation had no statistical significance. The average fixation duration showed a negative correlation with fixation points （r=-0.255, P<0.05）; the average fixation duration showed a negative correlation with average pupil diameter （r=-0.218, P<0.05）; the fixation points showed a positive correlation with average pupil diameter （r=0.09, P<0.05）. The area under the curve of average pupil diameter, average fixation duration and fixation points was 0.603, 0.621 and 0.580, respectively. Conclusion The average pupil diameter, average fixation duration and fixation points obtained by the eye tracker under laboratory conditions can be used to detect lies.
Algorithms ; Eye Movements ; Humans ; Lie Detection ; Pupil ; Time Factors

BACKGROUND/AIMS: We evaluated the effectiveness of an endoscopic ultrasonography (EUS)-based treatment plan compared to an endoscopy-based treatment plan in selecting candidates with early gastric cancer (EGC) for endoscopic submucosal dissection based on the prediction of invasion depth. METHODS: We reviewed 393 EGCs with differentiated histology from 380 patients who underwent EUS from July 2007 to April 2010. The effectiveness of the EUS-based and endoscopy-based plans was evaluated using a simplified hypothetical treatment algorithm. RESULTS: The numbers of endoscopically determined mucosal, indeterminate, and submucosal cancers were 253 (64.4%), 56 (14.2%), and 84 (21.4%), respectively. Overall, the appropriate treatment selection rates were 75.3% (296/393) in the endoscopy-based plan and 71.5% (281/393) in the EUS-based plan (p=0.184). For endoscopic mucosal cancers, the appropriate treatment selection rates in the endoscopy-based plan were 88.1% (223/253), while the use of an EUS-based plan significantly decreased this rate to 81.4% (206/253) (p=0.036). For endoscopic submucosal cancers, the appropriate selection rates did not differ between the endoscopy-based plan (46.4%, 39/84) and the EUS-based plan (53.6%, 45/84) (p=0.070). CONCLUSIONS: EUS did not increase the likelihood of selecting the appropriate treatment in differentiated-type EGC. Therefore, EUS may not be necessary before treating differentiated-type EGC, especially in endoscopically presumed mucosal cancers.
Adult ; Aged ; Aged, 80 and over ; Algorithms ; Clinical Decision-Making/*methods ; Early Detection of Cancer ; Endoscopy, Gastrointestinal/*statistics & numerical data ; Endosonography/*statistics & numerical data ; Female ; Gastric Mucosa/ultrasonography ; Humans ; Male ; Middle Aged ; *Patient Selection ; Prospective Studies ; Retrospective Studies ; Stomach Neoplasms/surgery/*ultrasonography

OBJECTIVES: This retrospective cohort study aimed to estimate the mean sojourn time (MST) of preclinical gastric cancer in Korean men. METHODS: The subjects consisted of voluntary male screenees aged 40 to 69 years who underwent subsequent screening gastroscopies after testing negative at a baseline screening performed between January 2007 and December 2011. A new case was defined if gastric cancer cells were present in the biopsy specimens obtained from gastroscopy. The follow-up period was calculated as the number of person-years between the date of baseline screening gastroscopy and positive findings at a subsequent screening. The MST was calculated using transition rates of gastric cancer to determine the best screening interval. RESULTS: Of the 171 979 voluntary male screenees, 61 688 (36%) underwent subsequent screening gastroscopies between January 2007 and December 2011. A total of 91 incident cases were found during 19 598 598 person-years of follow-up. The MST of gastric cancer was 2.37 years (95% confidence intervals, 1.92 to 2.96), and those aged 40 to 49 years had a shorter MST than those 50 to 69 years did. CONCLUSIONS: These findings support the 2-year interval of screening recommended by the nationwide gastric cancer screening program in Korea. Further studies for the age-specific MST among women are needed.
Adult ; Aged ; *Algorithms ; Asian Continental Ancestry Group ; Cohort Studies ; Early Detection of Cancer ; Gastroscopy ; Humans ; Incidence ; Male ; Mass Screening ; Middle Aged ; Republic of Korea ; Retrospective Studies ; Stomach Neoplasms/*epidemiology/pathology ; Time Factors

Algorithms ; Antigens, Neoplasm ; metabolism ; Carcinoembryonic Antigen ; metabolism ; Colorectal Neoplasms ; diagnosis ; Data Interpretation, Statistical ; Early Detection of Cancer ; Humans ; Mass Screening ; methods ; Middle Aged ; Pilot Projects ; Sensitivity and Specificity

OBJECTIVETo assess the feasibility and accuracy of different cervical cancer screening algorithms suitable for different regions, and promote the prevention and control of cervical cancer in China.

METHODSUsing the data of a cross-sectional comparative trial of multiple techniques to detect cervical intraepithelial neoplasia in Xiangyuan County, Shanxi Province, conducted in 1999, to evaluate the feasibility and accuracy of different cervical cancer screening algorithms. All the women were screened by six screening tests, including liquid based cytology (LBC), fluorescence spectroscopy, visual inspection with 5% acetic acid staining (VIA), colposcopy, self-sampled HPV DNA and clinician-sampled HPV DNA test, and each woman had histopathological diagnosis. Different screening algorithms were developed by combining the screening tests in parallel or in serial, and the performance indexes of the algorithms such as sensitivity, specificity, colposcopy referral rate and receiver operating characteristic (ROC) curve for detecting the high grade lesions (>or= CIN 2) were compared.

RESULTSAmong the algorithms combined by LBC and HPV DNA testing, for the combination in parallel (either cytology was greater than ASC-US or HPV positives), its sensitivity was 100%, specificity was 68.6%, and colposcopy referral rate was 34.4%; for the algorithm of LBC as primary screening test, with ASC-US women triage by HPV DNA testing, its sensitivity was 93.0%, specificity was 89.9%, and colposcopy referral rate was 13.7%; for the algorithm of HPV DNA testing as the primary screening test, with HPV positive women triage by LBC, its sensitivity was 91.7%, specificity was 93.0%, and colposcopy referral rate was 10.6%. ROC analysis showed that LBC primary testing followed by HPV triage and HPV primary testing followed by LBC triage were much better than the combination in parallel (P = 0.0003, P = 0.0002). Among the algorithms with LBC or HPV DNA testing solely as primary screening test, the sensitivity, specificity and colposcopy referral rate of LBC were 94.2%, 77.3%, 25.7% and 87.2%, 93.5%, 10.0%, respectively, for cutoff by ASC-US or by LSIL; the sensitivity, specificity and colposcopy referral rate of HPV DNA testing were 97.6%, 84.8%, 18.8% and 83.5%, 85.9%, 17.1%, respectively, for clinician-sampled and self-sampled. Clinician-sampled HPV DNA testing was better than LBC cutoff by ASC-US or self-sampled HPV DNA testing (P = 0.005, P = 0.002). Among the algorithms combined by VIA and HPV DNA testing, the sensitivity, specificity and colposcopy referral rate were 70.9%, 74.3% and 27.6% for VIA alone as primary screening test; the sensitivity, specificity and colposcopy referral rate were 65.9%, 95.2% and 7.4% for HPV as primary screening test with HPV positive women triage by VIA. HPV primary testing followed by VIA triage was better than VIA alone (P = 0.004).

CONCLUSIONConsidering the health-resource settings and women's preference, both HPV primary testing followed by LBC triage and LBC primary testing followed by HPV triage are suitable for developed regions, moderately developed regions may choose either LBC or HPV as the screening approach, VIA is a suitable primary screening test in less developed regions, and HPV primary testing followed by VIA triage will be more effective if low cost HPV test is available in the future.

Acetic Acid ; Adult ; Algorithms ; Cervical Intraepithelial Neoplasia ; diagnosis ; virology ; Colposcopy ; Cross-Sectional Studies ; Cytological Techniques ; methods ; DNA, Viral ; isolation & purification ; Early Detection of Cancer ; methods ; Feasibility Studies ; Female ; Humans ; Mass Screening ; methods ; Middle Aged ; Papillomaviridae ; isolation & purification ; Papillomavirus Infections ; diagnosis ; ROC Curve ; Staining and Labeling ; methods ; Uterine Cervical Neoplasms ; diagnosis ; virology

OBJECTIVE: To estimate the inevitable uncertainty of determining heroin by GC. METHODS: The source of the uncertainty was confirmed from the determining procedure. Each component's uncertainty was calculated. The combined uncertainty was then obtained by synthesizing the uncertainties of various components variables, and the expand uncertainty was finally obtained. RESULTS: The uncertainty of repeated measurement was greater than other uncertainty components introduced by balance,content vessels and instruments in determining heroin. CONCLUSION The errors of repeated measurement and GC instrumental were confirmed as the major sources of uncertainty in determining heroin by GC.
Algorithms ; Chromatography, Gas/methods* ; Forensic Medicine/methods* ; Heroin/chemistry* ; Humans ; Models, Theoretical ; Reproducibility of Results ; Solvents/chemistry* ; Substance Abuse Detection/methods* ; Uncertainty