Background removal for debiasing computer-aided cytological diagnosis.

To address the background-bias problem in computer-aided cytology caused by microscopic slide deterioration, this article proposes a deep learning approach for cell segmentation and background removal without requiring cell annotation. A U-Net-based model was trained to separate cells from the background in an unsupervised manner by leveraging the redundancy of the background and the sparsity of cells in liquid-based cytology (LBC) images. The experimental results demonstrate that the U-Net-based model trained on a small set of cytology images can exclude background features and accurately segment cells. This capability is beneficial for debiasing in the detection and classification of the cells of interest in oral LBC. Slide deterioration can significantly affect deep learning-based cell classification. Our proposed method effectively removes background features at no cost of cell annotation, thereby enabling accurate cytological diagnosis through the deep learning of microscopic slide images.

An effective salivation-facilitating method for COVID-19 testing: An experimental study.

BACKGROUND: Simple tools, such as antigen test kits, are readily available for determining coronavirus disease 2019 infection at hospitals and homes. However, it is challenging for elderly people who are prone to dry mouth and other diseases. The main objective of this study was to investigate whether the presence or consumption of a plum pickle can facilitate salivation during coronavirus disease 2019 testing. METHOD: Twenty healthy adult women participated in the study. The participants were allocated to 2 groups: presentation and non-presentation (n = 10; with and without presentation of a plum pickle, respectively), and eating and non-eating (n = 10; with and without consumption of plum pickle, respectively). We recorded the number of saliva swallows in 1 minute under each condition, using a swallowing test device, which attached film sensors to the hyoid bone and thyroid cartilage. RESULTS: There was a significant difference in the number of swallows between the non-presentation and presentation groups ( P < .01, r = 0.89, Z = -2.82) as well as between the non-eating and eating groups ( P < .01, r = 0.85, Z = -2.68). CONCLUSIONS: The strength of 3 factors, namely: direct stimulation with citric acid, saliva buffer capacity, and motor learning, may have affected the results. Our study suggests that saliva collection using the plum pickle is an effective complementary method for facilitating salivation. This technique may be useful in avoiding the risk associated with citric acid intake and for efficient specimen collection during coronavirus disease 2019 testing. In the future, we need to verify this method in elderly participants in a clinical setting.

Application of the sliding window method and Mask-RCNN method to nuclear recognition in oral cytology.

BACKGROUND: We aimed to develop an artificial intelligence (AI)-assisted oral cytology method, similar to cervical cytology. We focused on the detection of cell nuclei because the ratio of cell nuclei to cytoplasm increases with increasing cell malignancy. As an initial step in the development of AI-assisted cytology, we investigated two methods for the automatic detection of cell nuclei in blue-stained cells in cytopreparation images. METHODS: We evaluated the usefulness of the sliding window method (SWM) and mask region-based convolutional neural network (Mask-RCNN) in identifying the cell nuclei in oral cytopreparation images. Thirty cases of liquid-based oral cytology were analyzed. First, we performed the SWM by dividing each image into 96 × 96 pixels. Overall, 591 images with or without blue-stained cell nuclei were prepared as the training data and 197 as the test data (total: 1,576 images). Next, we performed the Mask-RCNN by preparing 130 images of Class II and III lesions and creating mask images showing cell regions based on these images. RESULTS: Using the SWM method, the highest detection rate for blue-stained cells in the evaluation group was 0.9314. For Mask-RCNN, 37 cell nuclei were identified, and 1 cell nucleus was identified as a non-nucleus after 40 epochs (error rate:0.027). CONCLUSIONS: Mask-RCNN is more accurate than SWM in identifying the cell nuclei. If the blue-stained cell nuclei can be correctly identified automatically, the entire cell morphology can be grasped faster, and the diagnostic performance of cytology can be improved.

Information-Theoretic Representation Learning for Positive-Unlabeled Classification.

Recent advances in weakly supervised classification allow us to train a classifier from only positive and unlabeled (PU) data. However, existing PU classification methods typically require an accurate estimate of the class-prior probability, a critical bottleneck particularly for high-dimensional data. This problem has been commonly addressed by applying principal component analysis in advance, but such unsupervised dimension reduction can collapse the underlying class structure. In this letter, we propose a novel representation learning method from PU data based on the information-maximization principle. Our method does not require class-prior estimation and thus can be used as a preprocessing method for PU classification. Through experiments, we demonstrate that our method, combined with deep neural networks, highly improves the accuracy of PU class-prior estimation, leading to state-of-the-art PU classification performance.

The acoustic characteristics of fine crackles predict honeycombing on high-resolution computed tomography.

BACKGROUND: Honeycombing on high-resolution computed tomography (HRCT) is a distinguishing feature of usual interstitial pneumonia and predictive of poor outcome in interstitial lung diseases (ILDs). Although fine crackles are common in ILD patients, the relationship between their acoustic features and honeycombing on HRCT has not been well characterized. METHODS: Lung sounds were digitally recorded from 71 patients with fine crackles and ILD findings on chest HRCT. Lung sounds were analyzed by fast Fourier analysis using a sound spectrometer (Easy-LSA; Fukuoka, Japan). The relationships between the acoustic features of fine crackles in inspiration phases (onset timing, number, frequency parameters, and time-expanded waveform parameters) and honeycombing in HRCT were investigated using multivariate logistic regression analysis. RESULTS: On analysis, the presence of honeycombing on HRCT was independently associated with onset timing (early vs. not early period; odds ratios [OR] 10.407, 95% confidence interval [95% CI] 1.366-79.298, P = 0.024), F99 value (the percentile frequency below which 99% of the total signal power is accumulated) (unit Hz = 100; OR 5.953, 95% CI 1.221-28.317, P = 0.029), and number of fine crackles in the inspiratory phase (unit number = 5; OR 4.256, 95% CI 1.098-16.507, P = 0.036). In the receiver-operating characteristic curves for number of crackles and F99 value, the cutoff levels for predicting the presence of honeycombing on HRCT were calculated as 13.2 (area under the curve [AUC], 0.913; sensitivity, 95.8%; specificity, 75.6%) and 752 Hz (AUC, 0.911; sensitivity, 91.7%; specificity, 85.2%), respectively. The multivariate logistic regression analysis additionally using these cutoff values revealed an independent association of number of fine crackles in the inspiratory phase, F99 value, and onset timing with the presence of honeycombing (OR 33.907, 95% CI 2.576-446.337, P = 0.007; OR 19.397, 95% CI 2.311-162.813, P = 0.006; and OR 12.383, 95% CI 1.443-106.293, P = 0.022; respectively). CONCLUSIONS: The acoustic properties of fine crackles distinguish the honeycombing from the non-honeycombing group. Furthermore, onset timing, number of crackles in the inspiratory phase, and F99 value of fine crackles were independently associated with the presence of honeycombing on HRCT. Thus, auscultation routinely performed in clinical settings combined with a respiratory sound analysis may be predictive of the presence of honeycombing on HRCT.

Development of a Skill Evaluation System for the Camera Assistant Using an Infant-Sized Laparoscopic Box Trainer.

AIMS: Our aims were to develop a training system for camera assistants (CA), and evaluate participants' performance as CA. METHODS: A questionnaire on essential requirements to be a good CA was administered to experts in pediatric endoscopic surgery. An infant-sized box trainer with several markers and lines inside was developed. Participants performed marker capturing and line-tracing tasks using a 5-mm 30° scope. A postexperimental questionnaire on the developed system was administered. The task completion time was measured. RESULTS: The 5-point evaluation scale was used for each item in the questionnaire survey of experts. The abilities to maintain a horizontal line (mean score: 4.5) and to center the target in a specified rectangle on the monitor (4.5) as well as having a full understanding of the operative procedure (4.3) were ranked as highly important. Fifty-two participants, including 5 surgical residents, were enrolled in the evaluation experiment. The completion time of capturing the markers was significantly longer in the resident group than in the nonresident group (244 versus 124 seconds, P = .04), but that of tracing the lines was not significantly different between the groups. The postexperimental questionnaire showed that the participants felt that the line-tracing tasks (3.7) were more difficult than marker-capturing tasks (2.9). CONCLUSIONS: Being proficient in manipulating a camera and having adequate knowledge of operative procedures are essential requirements to be a good CA. The ability was different between the resident and nonresident groups even in a simple task such as marker capturing.

Convex formulation of multiple instance learning from positive and unlabeled bags.

Multiple instance learning (MIL) is a variation of traditional supervised learning problems where data (referred to as bags) are composed of sub-elements (referred to as instances) and only bag labels are available. MIL has a variety of applications such as content-based image retrieval, text categorization, and medical diagnosis. Most of the previous work for MIL assume that training bags are fully labeled. However, it is often difficult to obtain an enough number of labeled bags in practical situations, while many unlabeled bags are available. A learning framework called PU classification (positive and unlabeled classification) can address this problem. In this paper, we propose a convex PU classification method to solve an MIL problem. We experimentally show that the proposed method achieves better performance with significantly lower computation costs than an existing method for PU-MIL.

Design and development of miniature parallel robot for eye surgery.

A five degree-of-freedom (DOF) miniature parallel robot has been developed to precisely and safely remove the thin internal limiting membrane in the eye ground during vitreoretinal surgery. A simulator has been developed to determine the design parameters of this robot. The developed robot's size is 85 mm × 100 mm × 240 mm, and its weight is 770 g. This robot incorporates an emergency instrument retraction function to quickly remove the instrument from the eye in case of sudden intraoperative complications such as bleeding. Experiments were conducted to evaluate the robot's performance in the master-slave configuration, and the results demonstrated that it had a tracing accuracy of 40.0 µm.

Reliability and repeatability of the Hoffmann sign.

OBJECTIVE: To determine the inter-rater, intrarater, and intrasubject reliability of the Hoffmann sign. DESIGN: Observational, cross-sectional study. SETTING: Veterans Affairs medical center. PATIENTS: Fifty-two consenting subjects without amputation of the first through third fingers, fixed finger contractures, relapsing remitting multiple sclerosis, or any acute central neurological illness or injury within the past 3 months requiring hospital admission were recruited from inpatients and outpatients in the Spinal Cord Injury and Physical Medicine and Rehabilitation services. INTERVENTIONS: The Hoffmann sign was elicited by 1 primary and 3 secondary investigators who used a standardized technique. The Hoffmann sign was considered positive if any reflexive flexion of the distal phalanx of the thumb or any of the fingers was present. In the first session, the primary and one secondary examiner performed 2 trials on both hands of each subject. Each investigator pair repeated the procedure in a second session. MAIN OUTCOME MEASURES: Cohen's κ coefficient was calculated to determine (1) inter-rater reliability, calculated per investigator pair per hand, for the first trial of a session; (2) intrarater reliability, calculated between the 2 trials of each session per investigator; and (3) intrasubject reliability, calculated per hand of each subject between the first trials of the 2 sessions. RESULTS: Inter-rater κ was 0.65 (188 pairs), intrarater κ was 0.89 (384 pairs), and intrasubject κ was 0.73 (178 pairs). All κ values were obtained with P < .01. CONCLUSIONS: The Hoffmann sign has substantial inter-rater and intrasubject reliability, in addition to outstanding intrarater reliability, when tested with the use of a standardized technique.

Mutagenesis of uracil-DNA glycosylase deficient mutants of the extremely thermophilic eubacterium Thermus thermophilus.

Thermus thermophilus is an extremely thermophilic, aerobic, and gram-negative eubacterium that grows optimally at 70-75 degrees C, pH 7.5. In extremely high temperature environment, DNA damages in cells occur at a much higher frequency in thermophiles than mesophiles such as E. coli. When temperature rises, the deamination of cytosine residues in double-strand DNA is expected to increase greatly. T. thermophilus HB27 has two putative uracil-DNA glycosylase genes (udgA and udgB). Expression level of udgA gene was 2-3 times higher than that of udgB at 70, 74, and 78 degrees C when it was monitored by beta-glucosidase reporter assay. We developed hisD(3110), hisD(3113), hisD(3115), and hisD(174) marker allele that can specifically detect G:C-->A:T, C:G-->A:T, T:A-->A:T, and A:T-->G:C base-substitutions, respectively, by His(+) reverse mutations. We then disrupted udgA and udgB by thermostable kanamycin-resistant gene (htk) or pyrE gene insertion in each hisD background, and their spontaneous His(+) reversion frequencies were compared. A udgA,B double mutant showed a pronounced increase in G:C-->A:T reversion frequency compared with each single udg mutant, udgA or udgB. Estimated mutation rates of the udgA,B mutant cultured at 60, 70, and 78 degrees C were about 2, 12, and 117 His(+)/10(8)/generation, respectively. At 70 degrees C culture, increased ratio of the mutation rate compared with the udg(+) strain was 12-fold in udgA, 3-fold in udgB, and 56-fold in udgA,B mutant. On the other hand, no difference was observed in other mutations of C:G-->A:T, T:A-->A:T, and A:T-->G:C between udgA,B double mutant and the parent udg(+) strain. The present results indicated that gene products of udgB as well as udgA functioned in vivo to remove uracil in DNA and prevent G:C-->A:T transition mutations.