DeepHealth Receives FDA Clearance for AI-Powered Breast Ultrasound

SOMERVILLE, Mass., July 30, 2026 (GLOBE NEWSWIRE) — DeepHealth, I c., a global leade i AI-powe ed health i fo matics a d a wholly ow ed subsidia y of RadNet, I c. (NASDAQ: RDNT), today a ou ced FDA 510(k) clea a ce fo DeepHealth B east Ult asou d,¹ a AI-powe ed solutio that b i gs g eate sta da dizatio , cli ical accu acy a d efficie cy to b east ult asou d imagi g. DeepHealth B east Ult asou d automates lesio detectio , cha acte izatio a d epo ti g, st eamli i g so og aphe a d adiologist wo kflows i o e i tellige t expe ie ce. Togethe with its AI-powe ed mammog aphy solutio s, the expa ded DeepHealth b east platfo m is the most comp ehe sive i the i dust y.

I additio to a multi- eade multi-case study i volvi g 16 U.S. boa d-ce tified adiologists at select imagi g ce te s a d hospitals,² RadNet has validated the effective ess of DeepHealth B east Ult asou d i live cli ical setti gs u de egulated esea ch p otocols. Now with FDA clea a ce, DeepHealth B east Ult asou d is comme cially available fo sale to custome s i the U ited States who ca pu sue eimbu seme t u de a existi g Catego y III CPT code fo qua titative ult asou d tissue cha acte izatio .³ By the e d of this yea , the solutio will be impleme ted ac oss RadNet’s etwo k of ce te s, with a estimated mo e tha 700,000 b east ult asou d studies a ually that may be eligible fo eimbu seme t.

“B east ult asou d is a esse tial compo e t of the b east ca e pathway, with app oximately 40% of wome u de goi g the exam at some poi t i thei lives.⁴ It is a highly complex, ope ato -depe de t exami atio , which ca lead to sig ifica t va iability i image acquisitio , i te p etatio a d epo ti g,” said D . Jaso McKellop, Medical Di ecto of Wome ‘s Imagi g fo RadNet Califo ia. “With DeepHealth’s b east ult asou d solutio , we ca achieve g eate sta da dizatio of wo kflows, imp ovi g co siste cy while savi g time fo patie ts, so og aphe s a d adiologists. By st eamli i g the exami atio p ocess, we ca help educe exam times, e ha ce efficie cy a d ultimately imp ove patie t outcomes.”

DeepHealth B east Ult asou d is desig ed to suppo t mo e efficie t a d sta da dized ult asou d wo kflows th ough:

Automated lesio detectio – Assists the i te p eti g physicia to localize the p ese ce o abse ce of suspicious soft-tissue lesio s i sta da d ult asou d b east images, with demo st ated g eate tha 98% accu acy i localizi g b east lesio s a d imp oved se sitivity fo b east ca ce detectio by 8%²
Automated lesio cha acte izatio – A alyzes images acqui ed by the so og aphe to assist the i te p eti g physicia i ge e ati g a d cha acte izi g the lesio shape, o ie tatio , ma gi , echo patte a d poste io featu es, i li e with ACR BI-RADS, educi g adiologist i te p etatio time by 37%²
Automated epo ti g – Ge e ates a comp ehe sive adiology epo t of the key fi di gs a d imp essio s, helpi g adiologists move mo e efficie tly f om image eview to fi al epo t while etai i g co t ol of the fi al assessme t
Expedited so og aphe wo kflow – Automatically ext acts a d o ga izes lesio measu eme ts, cha acte istics a d othe eleva t fi di gs i a sta da dized fo mat, educi g ma ual docume tatio fo so og aphe s with the aim of educi g ove all exam time

“No si gle imagi g pathway add esses eve y woma ’s eeds. With the additio of B east Ult asou d, we a e p oud to suppo t wome ac oss a b oade a ge of sc ee i g a d diag ostic pathways, i cludi g those with de se b easts a d othe s who may equi e suppleme tal imagi g. B i gi g togethe AI-powe ed capabilities ac oss mammog aphy a d ult asou d helps cli icia s espo d to diffe e t imagi g eeds a d delive mo e comp ehe sive, pe so alized b east ca e,” said Niccolò Stefa i, M.D., Busi ess a d P oduct Leade , Cli ical AI at DeepHealth.

DeepHealth’s b east imagi g platfo m⁵ — a modula , i te ope able AI-powe ed po tfolio add essi g eal-wo ld cli ical eeds ac oss b east ca ce sc ee i g a d diag ostic pathways — i cludes applicatio s fo ca ce detectio i both mammog aphy a d ult asou d, de sity assessme t, b east a te ial calcificatio assessme t, image-based b east ca ce isk p edictio ⁶ a d mammog aphy quality a alytics, with viewi g a d epo ti g tools fo imp oved ope atio al efficie cy.

About DeepHealthDeepHealth is a wholly ow ed subsidia y of RadNet, I c. (NASDAQ: RDNT) a d se ves as the umb ella b a d fo RadNet’s Digital Health segme t. DeepHealth p ovides AI-powe ed health i fo matics with the aim of empowe i g b eakth oughs i ca e th ough imagi g. DeepHealth leve ages adva ced AI fo ope atio al efficie cy a d imp oved cli ical outcomes i b east, chest, musculoskeletal, eu o, p ostate a d thy oid health. At the hea t of DeepHealth’s po tfolio is a cloud- ative ope ati g system – DeepHealth OS – that u ifies data ac oss the cli ical a d ope atio al wo kflow. Thousa ds of imagi g ce te s a d adiology depa tme ts a ou d the wo ld use DeepHealth solutio s to e able ea lie , mo e eliable a d mo e efficie t disease detectio , i cludi g i la ge-scale ca ce sc ee i g p og ams. DeepHealth’s huma -ce te ed, i tuitive tech ology aims to push the bou da ies of what’s possible i healthca e. Lea mo e at deephealth.com.

About RadNet, I c. RadNet, I c. is a leadi g atio al p ovide of f eesta di g, fixed-site diag ostic imagi g se vices i the U ited States based o the umbe of locatio s a d a ual imagi g eve ue. RadNet has a etwo k of ow ed a d/o ope ated outpatie t imagi g ce te s. RadNet’s imagi g ce te ma kets i clude A izo a, Califo ia, Delawa e, Flo ida, Idaho, I dia a, Ma yla d, New Je sey, New Yo k, Texas a d Vi gi ia. I additio , RadNet p ovides adiology i fo matio tech ology a d a tificial i tellige ce solutio s ma keted u de the DeepHealth b a d, tele adiology p ofessio al se vices a d othe elated p oducts a d se vices to custome s i the diag ostic imagi g i dust y globally. Togethe with co t acted adiologists, a d i clusive of full-time a d pe diem employees a d tech ologists, RadNet has ove 12,000 team membe s. Lea mo e at ad et.com.

Fo wa d Looki g Stateme tsThis commu icatio co tai s ce tai “fo wa d-looki g stateme ts” withi the mea i g of the safe ha bou p ovisio s of the U.S. P ivate Secu ities Litigatio Refo m Act of 1995, Sectio 27A of the Secu ities Act of 1933, as ame ded, a d Sectio 21E of the Secu ities Excha ge Act of 1934, as ame ded. Fo wa d-looki g stateme ts ca be ide tified by wo ds such as: “a ticipate,” “believe,” “could,” “estimate,” “expect,” “fo ecast,” “i te d,” “may,” “outlook,” “pla ,” “pote tial,” “possible,” “p edict,” “p oject,” “seek,” “should,” “ta get,” “will” o “would,” the egative of these wo ds, a d simila efe e ces to futu e pe iods. Examples of fo wa d-looki g stateme ts i clude stateme ts ega di g ou tech ology’s ability to stage-shift disease th ough p oactive, timely i te ve tio a d discussio s ega di g ou p oduct featu es. Actual esults could diffe mate ially f om those cu e tly a ticipated due to a umbe of isks a d u ce tai ties, ma y of which a e beyo d RadNet’s co t ol.

Fo wa d-looki g stateme ts a e eithe histo ical facts o assu a ces of futu e pe fo ma ce. I stead, they a e based o ly o ma ageme t’s cu e t beliefs, expectatio s a d assumptio s ega di g the futu e of RadNet’s busi ess, futu e pla s a d st ategies, p ojectio s, a ticipated eve ts a d t e ds, the eco omy a d othe futu e co ditio s. Because fo wa d-looki g stateme ts elate to the futu e, they a e subject to i he e t u ce tai ties, isks a d cha ges i ci cumsta ces that a e difficult to p edict a d ma y of which a e outside of RadNet’s co t ol. RadNet’s actual esults a d fi a cial co ditio may diffe mate ially f om those i dicated i the fo wa d-looki g stateme ts as a esult of va ious facto s. Neithe RadNet, o a y of its di ecto s, executive office s, o adviso s, p ovide a y ep ese tatio , assu a ce o gua a tee that the occu e ce of the eve ts exp essed o implied i a y fo wa d-looki g stateme ts will actually occu , o if a y of them do occu , what impact they will have o the busi ess, esults of ope atio s o fi a cial co ditio of RadNet. Should a y isks a d u ce tai ties develop i to actual eve ts, these developme ts could have a mate ial adve se effect o RadNet’s busi ess a d the ability to ealize the expected be efits of the tech ology. Risks a d u ce tai ties that could cause esults to diffe f om expectatio s i clude, but a e ot limited to: (1) the ability to ecog ize the a ticipated be efits of the tech ology, a d (2) the isk of legislative, egulato y, eco omic, competitive, a d tech ological cha ges, a d othe isks a d u ce tai ties desc ibed i the “Risk Facto s,” “Ma ageme t’s Discussio a d A alysis,” a d othe sectio s of ou fili gs with the Secu ities a d Excha ge Commissio , i cludi g ou most ece t A ual Repo t o Fo m 10-K a d Qua te ly Repo ts o Fo m 10-Q. The fo egoi g eview of impo ta t facto s should ot be co st ued as exhaustive a d should be ead i co ju ctio with the othe cautio a y stateme ts that a e i cluded elsewhe e. Additio al i fo matio co ce i g isks, u ce tai ties a d assumptio s ca be fou d i RadNet’s fili gs with the Secu ities a d Excha ge Commissio (the “SEC”), i cludi g the isk facto s discussed i RadNet’s most ece t A ual Repo t o Fo m 10-K, as updated by its Qua te ly Repo ts o Fo m 10-Q a d futu e fili gs with the SEC.

Fo wa d-looki g stateme ts i cluded he ei a e made o ly as of the date he eof a d, except as equi ed by applicable law, RadNet does ot u de take a y obligatio to update a y fo wa d-looki g stateme ts, o a y othe i fo matio i this commu icatio , as a esult of ew i fo matio , futu e developme ts o othe wise, o to co ect a y i accu acies o omissio s i them which become appa e t. All fo wa d-looki g stateme ts i this commu icatio a e qualified i thei e ti ety by this cautio a y stateme t.

DeepHealth Media Co tactA d a Axe teDi ecto of Commu icatio s+31614440971a d a.axe te@deephealth.com

RadNet Media Co tacts Ja e Mazu Se io Vice P eside t, Co po ate Commu icatio s +1 585-355-5978 ja e.mazu @ ad et.com

Ma k Stolpe Executive Vice P eside t a d Chief Fi a cial Office +1 310-445-2800

Refe e ces

Deephealth B east Ult asou d is the comme cial ame fo the See-Mode Augme ted Repo ti g Tool, B east (SMART-B) device, clea ed by the FDA u de 510(k) K260303. The clea a ce is held by See-Mode Tech ologies Pty Ltd, a DeepHealth compa y, which ma ufactu es the device; it is dist ibuted i the U ited States by DeepHealth, I c.
Data o file. Based o data f om MRMC study submitted to FDA fo 510(k) clea a ce K260303.
Reimbu seme t depe ds o paye policy.
Based o cha ge of egulatio s a d gold-sta da d app oach i b east ca ce sc ee i g. U.S. Food a d D ug Admi ist atio . Mammog aphy Quality Sta da ds Act a d Regulatio Ame dme ts: Small E tity Complia ce Guide—Guida ce fo I dust y a d Food a d D ug Admi ist atio Staff. Issued August 26, 2024, https://www.fda.gov/media/181152/dow load.U.S. Food a d D ug Admi ist atio . U de sta di g B east De sity. FDA Office of Wome ’s Health. https://www.fda.gov/co sume s/wome s-health-topics/u de sta di g-b east-de sity
DeepHealth b east solutio s comp ise multiple applicatio s, i cludi g Mammo Dx, B east De sity, Safegua d Review, Risk Assessme t, BAC Assessme t, Mammog aphy I sight, B east Ult asou d a d wo kflow tools i cludi g DeepHealth Viewe . DeepHealth Viewe is ma ufactu ed by eRAD, I c. a d dist ibuted by DeepHealth, I c. Mammog aphy I sights is ma ufactu ed by Aquila, I c. a d dist ibuted by DeepHealth, I c.
Not clea ed fo use i the U.S. Capability available i Eu ope.