Artificial intelligence-assisted CT characterizations and quantitative analysis for differentiating pre-invasive lesions from invasive adenocarcinomas in pulmonary subsolid nodules ≤ 2cm.

e21614 Background: The use of artificial intelligence (AI) in medical imaging has dramatically improved the quality of segmentation including accuracy, efficiency and reproducibility. This study sought to determine whether AI-assisted computed tomography (CT) features and quantitative analysis of pulmonary subsolid nodules (SSNs) under 2cm could be used to differentiate preinvasive lesions from invasive adenocarcinomas. Methods: Clinical data and CT images of 297 preinvasive lesions and early invasive lung adenocarcinomas confirmed by surgery pathology with CT manifestations of SSNs under 2cm were retrospectively analysed. The nodules were divided into two groups: the preinvasive lesions (PILs, N = 115) including 7 cases of atypical adenomatous hyperplasia (AAH), 30 cases of adenocarcinoma in situ (AIS) and 78 cases of minimally invasive adenocarcinoma (MIA), and the invasive adenocarcinomas (IACs, N = 182). All CTs were processed by AI and the volume, mean CT value, consolidation-to-tumor ratio (CTR), mass and maximum diameter of each SSN were obtained. Results: The volume, mean CT value, CTR, maximum diameter and mass of nodules showed significant difference between the two groups (Table). Multivariate analysis was determined by logistic regression. The regression model between the two groups was logit(p) = -1.439-2.927Volume +0.0005(mean CT value)-0.463(CTR > 0.5) +0.238(maximum diameter)+6.298(mass).The receiver operating characteristic curve (ROC) showed that the mass can do the best prediction among all the independent factors with the areas under the curve(AUC) 0.748 at a cut-off value of 0.154, with the sensitivity of 70.9% and specificity of 70.4% .The AUC of the ROC using the regression probabilities of regression model was 0.769. Conclusions: AI-assisted CT characterizations may be promising tools to predict if SSNs under 2 cm have invaded. [Table: see text]

Download Full-text

Identification of pathological subtypes of early lung adenocarcinoma based on artificial intelligence parameters and CT signs

Bioscience Reports ◽

10.1042/bsr20212416 ◽

2022 ◽

Author(s):

Weiyuan Fang ◽

Guorui Zhang ◽

Yali Yu ◽

Hongjie Chen ◽

Hong Liu

Keyword(s):

Artificial Intelligence ◽

Lung Adenocarcinoma ◽

Predictive Model ◽

Adenocarcinoma In Situ ◽

Invasive Adenocarcinoma ◽

Diagnostic Efficacy ◽

Ct Value ◽

Quantitative Parameters ◽

Mean Diameter ◽

Pathological Subtypes

Objective: To explore the value of quantitative parameters of artificial intelligence and computed tomography (CT) signs in identifying pathological subtypes of lung adenocarcinoma appearing as ground-glass nodules (GGNs). Methods: CT images of 224 GGNs from 210 individuals were collected retrospectively and pathologically classified into atypical adenomatous hyperplasia (AAH)/adenocarcinoma in situ (AIS), minimally invasive adenocarcinoma (MIA), and invasive adenocarcinoma (IAC) groups. Artificial intelligence was used to identify GGNs and to obtain quantitative parameters, and CT signs were recognized manually. The mixed predictive model based on logistic multivariate regression was evaluated. Results: Of the 224 GGNs, 55, 93, and 76 were AAH/AIS, MIA, IAC, respectively. In terms of artificial intelligence parameters, from AAH/AIS to MIA, and IAC, there was a gradual increase in two-dimensional mean diameter, three-dimensional mean diameter, mean CT value, maximum CT value, and volume of GGNs (all P < 0.0001). Except for the CT signs of the location, and the tumor-lung interface, there were significant differences among the three groups in the density type, shape, vacuole signs, air bronchogram, lobulation, spiculation, pleural indentation, and vascular convergence signs (all P < 0.05). The areas under the curve (AUC) of predictive model 1 for identifying the AAH/AIS and MIA and model 2 for identifying MIA and IAC were 0.779 and 0.918, respectively, which were greater than the quantitative parameters independently (all P < 0.05). Conclusion: Artificial intelligence parameters are valuable for identifying subtypes of early lung adenocarcinoma, and when combined with CT signs to improve its diagnostic efficacy.

Download Full-text

Substation Decision-making Platform Based on Artificial Intelligence

Distributed Generation & Alternative Energy Journal ◽

10.13052/dgaej2156-3306.3524 ◽

2021 ◽

Author(s):

Jianhua Qin ◽

Xueqiong Zhu ◽

Zhen Wang ◽

Jingtan Ma ◽

Shan Gao ◽

...

Keyword(s):

Artificial Intelligence ◽

Decision Making ◽

Logistic Regression ◽

Quantitative Analysis ◽

Regression Model ◽

Logistic Regression Model ◽

Final Decision ◽

Equipment Failure ◽

Operational Data ◽

Electrical Data

In view of the actual needs faced by the substation maintenance, this paper proposes a kind of substation decision-making platform based on artificial intelligence. The platform formalizes and integrates the basic data, electrical data and the operational data of the equipment, qualitatively triggers the maintenance task abide by the result of the logistic regression model, provides further results of data processing through quantitative analysis, and provides knowledge navigation to the operation guidance of the corresponding equipment. The platform matches the electrical data with the inference engine stored in the knowledge base. If the data match the condition of the inference successfully, the inference is triggered and the action is executed. The result is provided to the relevant staff as a suggestion to assist the final decision. After the task is completed, the cause, effect and solution of the equipment failure are backfilled and expanded into the equipment base as a new instance.

Download Full-text

Adenocarcinoma in situ and minimally invasive adenocarcinoma in lungs of smokers: image feature differences from those in lungs of non-smokers

BMC Medical Imaging ◽

10.1186/s12880-021-00705-1 ◽

2021 ◽

Vol 21 (1) ◽

Author(s):

Haruto Sugawara ◽

Hirokazu Watanabe ◽

Akira Kunimatsu ◽

Osamu Abe ◽

Shun-ichi Watanabe ◽

...

Keyword(s):

Minimally Invasive ◽

Ct Images ◽

Solid Component ◽

Image Feature ◽

Adenocarcinoma In Situ ◽

Invasive Adenocarcinoma ◽

Invasive Component ◽

Significant Difference ◽

Minimally Invasive Adenocarcinoma

Abstract Purpose We aimed to examine the characteristics of imaging findings of adenocarcinoma in situ (AIS) and minimally invasive adenocarcinoma (MIA) in the lungs of smokers compared with those of non-smokers. Materials and methods We included seven cases of AIS and 20 cases of MIA in lungs of smokers (pack-years ≥ 20) and the same number of cases of AIS and MIA in lungs of non-smokers (pack-years = 0). We compared the diameter of the entire lesion and solid component measured on computed tomography (CT) images, pathological size and invasive component diameter measured from pathological specimens, and CT values of the entire lesion and ground-glass opacity (GGO) portions between the smoker and non-smoker groups. Results The diameters of AIS and MIA on CT images and pathological specimens of the smoker group were significantly larger than those of the non-smoker group (p = 0.036 and 0.008, respectively), whereas there was no significant difference in the diameter of the solid component on CT images or invasive component of pathological specimens between the two groups. Additionally, mean CT values of the entire lesion and GGO component of the lesions in the smoker group were significantly lower than those in the non-smoker group (p = 0.036 and 0.040, respectively). Conclusion AIS and MIA in smoker’s lung tended to have larger lesion diameter and lower internal CT values compared with lesions in non-smoker’s lung. This study calls an attention on smoking status in CT-based diagnosis for early stage adenocarcinoma.

Download Full-text

Lung Adenocarcinoma Invasiveness Risk in Pure Ground-Glass Opacity Lung Nodules Smaller than 2 cm

The Thoracic and Cardiovascular Surgeon ◽

10.1055/s-0037-1612615 ◽

2018 ◽

Vol 67 (04) ◽

pp. 321-328 ◽

Cited By ~ 8

Author(s):

Geun Dong Lee ◽

Chul Hwan Park ◽

Heae Surng Park ◽

Min Kwang Byun ◽

Ik Jae Lee ◽

...

Keyword(s):

Lung Cancer ◽

Lung Adenocarcinoma ◽

Wedge Resection ◽

Ground Glass Opacity ◽

Ground Glass ◽

Maximum Diameter ◽

Lung Nodules ◽

Adenocarcinoma In Situ ◽

Invasive Adenocarcinoma ◽

Maximal Diameter

Background We aimed to identify clinicopathologic characteristics and risk of invasiveness of lung adenocarcinoma in surgically resected pure ground-glass opacity lung nodules (GGNs) smaller than 2 cm. Methods Among 755 operations for lung cancer or tumors suspicious for lung cancer performed from 2012 to 2016, we retrospectively analyzed 44 surgically resected pure GGNs smaller than 2 cm in diameter on computed tomography (CT). Results The study group was composed of 36 patients including 11 men and 25 women with a median age of 59.5 years (range, 34–77). Median follow-up duration of pure GGNs was 6 months (range, 0–63). Median maximum diameter of pure GGNs was 8.5 mm (range, 4–19). Pure GGNs were resected by wedge resection, segmentectomy, or lobectomy in 27 (61.4%), 10 (22.7%), and 7 (15.9%) cases, respectively. Pathologic diagnosis was atypical adenomatous hyperplasia, adenocarcinoma in situ, minimally invasive adenocarcinoma (MIA), or invasive adenocarcinoma (IA) in 1 (2.3%), 18 (40.9%), 15 (34.1%), and 10 (22.7%) cases, respectively. The optimal cutoff value for CT-maximal diameter to predict MIA or IA was 9.1 mm. In multivariate analyses, maximal CT-maximal diameter of GGNs ≥10 mm (odds ratio, 24.050; 95% confidence interval, 2.6–221.908; p = 0.005) emerged as significant independent predictor for either MIA or IA. Estimated risks of MIA or IA were 37.2, 59.3, 78.2, and 89.8% at maximal GGN diameters of 5, 10, 15, and 20 mm, respectively. Conclusion Pure GGNs were highly associated with lung adenocarcinoma in surgically resected cases, while estimated risk of GGNs invasiveness gradually increased as maximal diameter increased.

Download Full-text

CT features and artificial intelligence quantitative analysis of recovered COVID-19 patients with negative RT-PCR and clinical symptoms

10.21203/rs.3.rs-21021/v1 ◽

2020 ◽

Author(s):

Siyao Du ◽

Si Gao ◽

Guoliang Huang ◽

Shu Li ◽

Wei Chong ◽

...

Keyword(s):

Artificial Intelligence ◽

Quantitative Analysis ◽

Clinical Symptoms ◽

Lower Lobe ◽

Ground Glass Opacity ◽

Imaging Features ◽

Rt Pcr ◽

Ct Value ◽

Novel Coronavirus

Abstract Objectives: To evaluate imaging features and performed quantitative analysis for mild novel coronavirus pneumonia (COVID-19) cases ready for discharge.Methods: CT images of 125 patients (16-67 years, 63 males) recovering from COVID-19 were examined. We defined the double-negative period (DNp) as the period between the sampling days of two consecutive negative RT-PCR and three days thereafter. Lesion demonstrations and distributions on CT in DNp (CTDN) were evaluated by radiologists and artificial intelligence (AI) software. Major lesion transformations and the involvement range for patients with follow-up CT were analyzed.Results: Twenty (16.0%) patients exhibited normal CTDN; abnormal CTDN for 105 indicated ground-glass opacity (GGO) (99/125, 79.2%) and fibrosis (56/125, 44.8%) as the most frequent CT findings. Bilateral-lung involvement with mixed or random distribution was most common for GGO on CTDN. Fibrous lesions often affected both lungs, tending to distribute on the subpleura. Follow-up CT showed lesion improvement manifesting as GGO thinning (40/40, 100%), fibrosis reduction (17/26, 65.4%), and consolidation fading (9/11, 81.8%), with or without range reduction. AI analysis showed the highest proportions for right lower lobe involvement (volume, 12.01±35.87cm3; percentage; 1.45±4.58%) and CT-value ranging –570 to –470 HU (volume, 2.93±7.04cm3; percentage, 5.28±6.47%). Among cases with follow-up CT, most of lung lobes and CT-value ranges displayed a significant reduction after DNp.Conclusions: The main CT imaging manifestations were GGO and fibrosis in DNp, which weakened with or without volume reduction. AI analysis results were consistent with imaging features and changes, possibly serving as an objective indicator for disease monitoring and discharge.

Download Full-text

Logistic Regression Model to Distinguish Between the Benign and Malignant Adnexal Mass Before Surgery: A Multicenter Study by the International Ovarian Tumor Analysis Group

Journal of Clinical Oncology ◽

10.1200/jco.2005.01.7632 ◽

2005 ◽

Vol 23 (34) ◽

pp. 8794-8801 ◽

Cited By ~ 274

Author(s):

Dirk Timmerman ◽

Antonia C. Testa ◽

Tom Bourne ◽

Enrico Ferrazzi ◽

Lieveke Ameye ◽

...

Keyword(s):

Blood Flow ◽

Logistic Regression ◽

Regression Model ◽

Logistic Regression Model ◽

Malignant Tumors ◽

Characteristic Curve ◽

Maximum Diameter ◽

Benign Tumors ◽

Data Set ◽

Maximal Diameter

Purpose To collect data for the development of a more universally useful logistic regression model to distinguish between a malignant and benign adnexal tumor before surgery. Patients and Methods Patients had at least one persistent mass. More than 50 clinical and sonographic end points were defined and recorded for analysis. The outcome measure was the histologic classification of excised tissues as malignant or benign. Results Data from 1,066 patients recruited from nine European centers were included in the analysis; 800 patients (75%) had benign tumors and 266 (25%) had malignant tumors. The most useful independent prognostic variables for the logistic regression model were as follows: (1) personal history of ovarian cancer, (2) hormonal therapy, (3) age, (4) maximum diameter of lesion, (5) pain, (6) ascites, (7) blood flow within a solid papillary projection, (8) presence of an entirely solid tumor, (9) maximal diameter of solid component, (10) irregular internal cyst walls, (11) acoustic shadows, and (12) a color score of intratumoral blood flow. The model containing all 12 variables (M1) gave an area under the receiver operating characteristic curve of 0.95 for the development data set (n = 754 patients). The corresponding value for the test data set (n = 312 patients) was 0.94; and a probability cutoff value of .10 gave a sensitivity of 93% and a specificity of 76%. Conclusion Because the model was constructed from multicenter data, it is more likely to be generally applicable. The effectiveness of the model will be tested prospectively at different centers.

Download Full-text

Artificial Intelligence for Skin Cancer Detection: A Scoping Review (Preprint)

10.2196/preprints.22934 ◽

2020 ◽

Author(s):

Abdulrahman Takiddin ◽

Jens Schneider ◽

Yin Yang ◽

Alaa Abd-Alrazaq ◽

Mowafa Househ

Keyword(s):

Artificial Intelligence ◽

Skin Cancer ◽

Scoping Review ◽

Performance Metrics ◽

Data Extraction ◽

Characteristic Curve ◽

Evaluation Metrics ◽

Cancer Type ◽

Computer Algorithms ◽

Scoping Reviews

BACKGROUND Skin cancer is the most common cancer type affecting humans. Traditional skin cancer diagnosis methods are costly, require a professional physician, and take time. Hence, to aid in diagnosing skin cancer, Artificial Intelligence (AI) tools are being used, including shallow and deep machine learning-based techniques that are trained to detect and classify skin cancer using computer algorithms and deep neural networks. OBJECTIVE The aim of this study is to identify and group the different types of AI-based technologies used to detect and classify skin cancer. The study also examines the reliability of the selected papers by studying the correlation between the dataset size and number of diagnostic classes with the performance metrics used to evaluate the models. METHODS We conducted a systematic search for articles using IEEE Xplore, ACM DL, and Ovid MEDLINE databases following the PRISMA Extension for Scoping Reviews (PRISMA-ScR) guidelines. The study included in this scoping review had to fulfill several selection criteria; to be specifically about skin cancer, detecting or classifying skin cancer, and using AI technologies. Study selection and data extraction were conducted by two reviewers independently. Extracted data were synthesized narratively, where studies were grouped based on the diagnostic AI techniques and their evaluation metrics. RESULTS We retrieved 906 papers from the 3 databases, but 53 studies were eligible for this review. While shallow techniques were used in 14 studies, deep techniques were utilized in 39 studies. The studies used accuracy (n=43/53), the area under receiver operating characteristic curve (n=5/53), sensitivity (n=3/53), and F1-score (n=2/53) to assess the proposed models. Studies that use smaller datasets and fewer diagnostic classes tend to have higher reported accuracy scores. CONCLUSIONS The adaptation of AI in the medical field facilitates the diagnosis process of skin cancer. However, the reliability of most AI tools is questionable since small datasets or low numbers of diagnostic classes are used. In addition, a direct comparison between methods is hindered by a varied use of different evaluation metrics and image types.

Download Full-text

Artificial intelligence-based education assists medical students’ interpretation of hip fracture

Insights into Imaging ◽

10.1186/s13244-020-00932-0 ◽

2020 ◽

Vol 11 (1) ◽

Author(s):

Chi-Tung Cheng ◽

Chih-Chi Chen ◽

Chih-Yuan Fu ◽

Chung-Hsien Chaou ◽

Yu-Tung Wu ◽

...

Keyword(s):

Artificial Intelligence ◽

Medical Education ◽

Hip Fracture ◽

Medical Students ◽

Medical Image ◽

Learning System ◽

Significant Difference ◽

Dynamic Collaboration ◽

Increase In Accuracy ◽

Image Learning

Abstract Background With recent transformations in medical education, the integration of technology to improve medical students’ abilities has become feasible. Artificial intelligence (AI) has impacted several aspects of healthcare. However, few studies have focused on medical education. We performed an AI-assisted education study and confirmed that AI can accelerate trainees’ medical image learning. Materials We developed an AI-based medical image learning system to highlight hip fracture on a plain pelvic film. Thirty medical students were divided into a conventional (CL) group and an AI-assisted learning (AIL) group. In the CL group, the participants received a prelearning test and a postlearning test. In the AIL group, the participants received another test with AI-assisted education before the postlearning test. Then, we analyzed changes in diagnostic accuracy. Results The prelearning performance was comparable in both groups. In the CL group, postlearning accuracy (78.66 ± 14.53) was higher than prelearning accuracy (75.86 ± 11.36) with no significant difference (p = .264). The AIL group showed remarkable improvement. The WithAI score (88.87 ± 5.51) was significantly higher than the prelearning score (75.73 ± 10.58, p < 0.01). Moreover, the postlearning score (84.93 ± 14.53) was better than the prelearning score (p < 0.01). The increase in accuracy was significantly higher in the AIL group than in the CL group. Conclusion The study demonstrated the viability of AI for augmenting medical education. Integrating AI into medical education requires dynamic collaboration from research, clinical, and educational perspectives.

Download Full-text

Incidental Gallbladder Cancer: Routine versus Selective Histological Examination After Cholecystectomy

The Surgery Journal ◽

10.1055/s-0040-1722175 ◽

2021 ◽

Vol 07 (01) ◽

pp. e22-e25

Author(s):

Andrew Alabi ◽

A D. Arvind ◽

Nikhil Pawa ◽

Shakir Karim ◽

Jason Smith

Keyword(s):

Gallbladder Cancer ◽

Gallstone Disease ◽

Papillary Adenocarcinoma ◽

University Teaching ◽

Histological Evaluation ◽

Adenocarcinoma In Situ ◽

Invasive Adenocarcinoma ◽

Potential Cost ◽

Histopathological Evaluation ◽

Incidental Gallbladder Cancer

Abstract Background Incidental gallbladder cancer is relatively rare, with an incidence ranging between 0.19 and 5.5% of all the cholecystectomies for benign disease, and carries a poor prognosis. Currently, in the literature, there appears to be some controversy about whether all gallbladder specimens should be sent for routine histopathology. The aim of this study was to investigate the need for either routine or selective histopathological evaluation of all gallbladder specimens following cholecystectomy in our institution. Methods The records of all patients who underwent a cholecystectomy (laparoscopic and open) for gallstone disease over a 5-year period (between January 2011 and January 2016) were reviewed retrospectively in a single university teaching hospital. Patients with radiological evidence of gallbladder cancer preoperatively were excluded. The notes of patients with incidental gallbladder cancer were reviewed and data were collected for clinical presentation and preoperative investigations including blood tests and radiological imaging. Results A total of 1,473 specimens were sent for histopathological evaluation, with two patients being diagnosed with an incidental gallbladder cancer (papillary adenocarcinoma in situ and moderately differentiated invasive adenocarcinoma [stage IIIa]). The incidence rate was 0.14%. All patients with incidental gallbladder cancer had macroscopically abnormal specimens. Conclusion Both patients in our study who were diagnosed with incidental gallbladder cancer had macroscopic abnormalities. A selective rather than routine approach to histological evaluation of gallbladder specimens especially in those with macroscopic abnormalities should be employed. This will reduce the burden on the pathology department with potential cost savings.

Download Full-text

Stress testing reveals gaps in clinic readiness of image-based diagnostic artificial intelligence models

npj Digital Medicine ◽

10.1038/s41746-020-00380-6 ◽

2021 ◽

Vol 4 (1) ◽

Author(s):

Albert T. Young ◽

Kristen Fernandez ◽

Jacob Pfau ◽

Rasika Reddy ◽

Nhat Anh Cao ◽

...

Keyword(s):

Artificial Intelligence ◽

Real World ◽

Stress Testing ◽

Operating Characteristic ◽

Characteristic Curve ◽

Skin Lesions ◽

Stress Tests ◽

Performance Metric ◽

Intelligence Models ◽

Artificial Intelligence Models

AbstractArtificial intelligence models match or exceed dermatologists in melanoma image classification. Less is known about their robustness against real-world variations, and clinicians may incorrectly assume that a model with an acceptable area under the receiver operating characteristic curve or related performance metric is ready for clinical use. Here, we systematically assessed the performance of dermatologist-level convolutional neural networks (CNNs) on real-world non-curated images by applying computational “stress tests”. Our goal was to create a proxy environment in which to comprehensively test the generalizability of off-the-shelf CNNs developed without training or evaluation protocols specific to individual clinics. We found inconsistent predictions on images captured repeatedly in the same setting or subjected to simple transformations (e.g., rotation). Such transformations resulted in false positive or negative predictions for 6.5–22% of skin lesions across test datasets. Our findings indicate that models meeting conventionally reported metrics need further validation with computational stress tests to assess clinic readiness.

Download Full-text