Kvíz spojený s průvodcem · Těžko úroveň
Kvíz AI v mamografii a screeningu rakoviny prsu
Testuje AI jako druhý nástroj pro čtení a třídění, co zjistily velké screeningové studie jako MASAI, falešně pozitivní kompromisy a limity husté prsní tkáně.
Otázka 1 z 8
Jak švédská studie MASAI použila AI při screeningu?
Náhled každé otázky8 otázky
Odpovědi a vysvětlení se zobrazují, jak budete v kvízu.
- Jak švédská studie MASAI použila AI při screeningu?
- AI read all exams with no radiologists involved
- AI was used only to schedule appointments
- AI triaged exams so low-risk ones got one reader and high-risk ones got two
- AI replaced mammograms with ultrasound
- Co zjistily velké studie o tradičním CAD představeném na konci 90. let?
- It did not clearly improve accuracy and added false marks
- It eliminated false positives
- It doubled survival rates
- It was never used in the United States
- Proč hustá prsní tkáň ztěžuje screening?
- Dense tissue blocks the X-ray machine from working
- Dense tissue appears white on mammograms, like many tumors, and can hide cancers
- Dense tissue only occurs in men
- Dense tissue makes images too dark
- Proč není odhalení většího počtu rakovin samo o sobě důkazem, že umělá inteligence zachraňuje životy?
- Detection rates are never measured
- All detected cancers are harmless
- Mammograms cannot detect cancer
- Some detected cancers grow slowly, so interval and advanced cancer rates are key long-term measures
- Co se stane, když program sníží provozní práh systému AI?
- It catches more cancers but increases recalls
- It reduces both detections and recalls
- It has no effect on results
- It makes images higher resolution
- What is automation bias in AI-supported mammography?
- AI becoming faster over time
- Machines producing biased images
- Readers overtrusting AI markers or skipping areas the AI did not flag
- Radiologists refusing all AI tools
- Why should screening programs validate AI locally even after regulatory clearance?
- Clearance expires after one day
- Performance can shift with different machines, image processing and populations
- Local validation makes the AI learn new languages
- Regulators require AI to be retrained daily
- In current AI-supported screening programs, who makes recall decisions?
- The AI system alone
- Pacient
- The equipment manufacturer
- Radiologists