Evidence map›Paper›PMID 37477673›Full record

ArticleEuropean radiology2023

Efficacy of exponentiation method with a convolutional neural network for classifying lung nodules on CT images by malignancy level.

Takuma Usuzaki, Kengo Takahashi, Hidenobu Takagi, Mami Ishikuro, Taku Obara, Takumi Yamaura, Masahiro Kamimoto, Kazuhiro Majima

Abstract read
PubMed Publisher
In one paragraph

Article in European radiology, 2023. The graph could read no effect estimate from its abstract, so it casts no vote on the map. Cited by 2 papers.

0numbers the graph read from it
0cells of the map it votes in
2citing papers in PubMed
–field-weighted citation impact
1 · What the graph read from it

What it found

Each row is one number read from the abstract, on the scale the paper reported it, with its interval. Left of the dashed line favours the treatment, right favours the comparator. Under each row is the sentence it came from. New to these charts? A ten-minute tutorial.

The abstract states no effect estimate the extractor could read, or names no intervention and outcome on the map, so this paper lights no cell and moves no belief. It is still indexed, cited and linked below.

2 · The registry

The trial behind it

Trials whose registry record cites this paper, or whose number appears in the abstract. A trial that started after this paper was published is citing it as background, not reporting it.

Neither the registry nor the abstract names a trial number. If this is a trial report, that itself is worth knowing.

3 · Its place in the literature

Who cites it

2 citing papers in PubMed.

  1. Article
  2. Article
4 · The record

Corrections and comments

PubMed lists nothing against this paper. Absence here is not a guarantee, only a check that was made.

5 · Who and what money

Authors and funding

8 authors.

Takuma UsuzakiDepartment of Diagnostic Radiology, Tohoku University Hospital, 1-1 Seiryo-Machi, Aoba-Ku, Sendai, Miyagi, 980-8574, Japan. takuma.usuzaki.p6@dc.tohoku.ac.jp.ORCID http://orcid.org/0000-0002-3945-3432
Kengo TakahashiTohoku University Graduate School of Medicine, Sendai, Japan.
Hidenobu TakagiDepartment of Diagnostic Radiology, Tohoku University Hospital, 1-1 Seiryo-Machi, Aoba-Ku, Sendai, Miyagi, 980-8574, Japan.
Mami IshikuroDivision of Molecular Epidemiology, Graduate School of Medicine, Tohoku University, Sendai, Miyagi, Japan.
Taku ObaraDivision of Molecular Epidemiology, Graduate School of Medicine, Tohoku University, Sendai, Miyagi, Japan.
Takumi YamauraTakeda General Hospital, Aizuwakamatsu, Japan.
Masahiro KamimotoTakeda General Hospital, Aizuwakamatsu, Japan.
Kazuhiro MajimaTakeda General Hospital, Aizuwakamatsu, Japan.

Funding

No grant is acknowledged in the PubMed record.

6 · The paper itself

Abstract

objectivesThe aim of this study was to examine the performance of a convolutional neural network (CNN) combined with exponentiating each pixel value in classifying benign and malignant lung nodules on computed tomography (CT) images. MATERIALS AND

methodsImages in the Lung Image Database Consortium-Image Database Resource Initiative (LIDC-IDRI) were analyzed. Four CNN models were then constructed to classify the lung nodules by malignancy level (malignancy level 1 vs. 2, malignancy level 1 vs. 3, malignancy level 1 vs. 4, and malignancy level 1 vs. 5). The exponentiation method was applied for exponent values of 1.0 to 10.0 in increments of 0.5. Accuracy, sensitivity, specificity, and area under the curve of receiver operating characteristics (AUC-ROC) were calculated. These statistics were compared between an exponent value of 1.0 and all other exponent values in each model by the Mann-Whitney U-test.

resultsIn malignancy 1 vs. 4, maximum test accuracy (MTA; exponent value = 2.0, 3.0, 3.5, 4.5, 5.0, 5.5, 6.0, 6.5, 7.0, 7.5, 8.0, 8.5, 9.0, 9.5, and 10.0) and specificity (6.5, 7.0, and 9.0) were improved by up to 0.012 and 0.037, respectively. In malignancy 1 vs. 5, MTA (6.5 and 7.0) and sensitivity (1.5) were improved by up to 0.030 and 0.0040, respectively.

conclusionsThe exponentiation method improved the performance of the CNN in the task of classifying lung nodules on CT images as benign or malignant. The exponentiation method demonstrated two advantages: improved accuracy, and the ability to adjust sensitivity and specificity by selecting an appropriate exponent value. CLINICAL RELEVANCE STATEMENT: Adjustment of sensitivity and specificity by selecting an exponent value enables the construction of proper CNN models for screening, diagnosis, and treatment processes among patients with lung nodules. KEY POINTS: • The exponentiation method improved the performance of the convolutional neural network. • Contrast accentuation by the exponentiation method may derive features of lung nodules. • Sensitivity and specificity can be adjusted by selecting an exponent value.

Indexed as

Lung NeoplasmsSolitary Pulmonary NoduleHumansLungNeural Networks, ComputerROC CurveTomography, X-Ray ComputedArtificial intelligenceDeep learningLung neoplasmsNeural networksSupervised machine learning

Identifiers

PMID37477673

What OpenQuestion holds

Textmetadata
Read underepoch 390

Registered trials

None linked

Read under generation 80e0d062 · epoch 390. Bibliography from PubMed, PubMed Central and OpenAlex; grants from NIH RePORTER; trial links from ClinicalTrials.gov; estimates, votes and beliefs from the OpenQuestion graph.