orla guerin, bbc news, port—au—prince. this is bbc news. we'll have the headlines and all the other main news stories for you at the top of the hour, straight after this programme. this week's programme is a mix of medical marvels. we'll look at the cutting—edge gene therapy that could bring hope to millions. we'll find out how artificial intelligence is helping doctors assess the risk of breast cancer returning.

and see the smartglasses helping make sense of the world. artificial voice: find object. find people. explore. plus zoe breaks down 56, and everything else in her vicinity. oh, god, that is really heavy! wow. what a beautiful building. this is europe's biggest biomedical research facility under one roof. it's the francis crick institute in london. there are over 1,500 scientists working here, all looking at the biology that underlies health and disease. whether it's cancer, heart disease or strokes, the research here hopes to change the way that serious illness is diagnosed, treated, or even prevented. the institute gets its name from uk scientist francis crick, one of the four scientists who discovered this — the shape of our dna — which in turn determines the shape of us.

there is dna in the nucleus of every cell. it carries genetic information with all the instructions that a living organism needs to grow, reproduce and function. genes are short sections of dna which carry information about different characteristics like ear shape, eye colour. they've even got one for my poor sense of humour. he isjoking. my point, exactly. but sometimes they can go wrong, and that can lead to genetic disorders or disease. however, scientists are getting better and better at fixing genes. it's called gene therapy, and paul carter has been to visit one company at the cutting edge. paul: modern medicine continues to develop at a rapid pace, and advances in science and technology continue to drive innovation. this is the downstream processing labs. in london, i've been invited behind the scenes to see some pioneering medical research

that has the potential to change the lives of people with inherited conditions or diseases. meiragtx is a company developing gene—therapy treatments. while gene therapy itself has existed for around two decades, scientists and researchers here are looking to progress it to the next level. traditional gene therapy works by replacing broken or faulty genes that are causing a condition with working copies of those genes. the therapies being developed here use harmless viruses to carry the genes, which are being paired with a form of biological switch that can be taken in tablet form. where disease is caused by a gene becoming incorrectly switched on, this treatment tries to turn it off — or vice versa. when the idea of gene therapy

started, one of the big issues was how do you get those genes into people? so we, and others, over the last five to ten years, have developed a new viral vector, and we're manufacturing that here to deliver the genes we're delivering into the body. in addition, at meira, we've actually developed, over the last seven years, a totally new technology which allows us to, in the future, control how those genes delivered to the body are activated using pills. one of the six clinical trials taking place here is targeting an inherited eye condition called retinitis pigmentosa. it causes light—detecting cells in the retina to break down over time, causing a person to eventually lose vision. red are the rod cells and green are the cone cells, so they exist in this kind of mosaic. early results of using gene

therapy to treat retinitis pigmentosa has shown that it could potentially not only slow the progression of the disease, but in some cases reverse some of its effects. the work being done here has real potential to change the lives of people with inherited conditions. the challenge is taking it out of a laboratory like this and getting it to those people that really need it. the retinitis pigmentosa research is being carried out with pharmaceutical companyjohnson & johnson and is currently in stage three clinical trials. it's expected that approval will be filed for in 2024, meaning that we could be seeing treatment available in years rather than in decades. this kind of gene therapy is showing promise in other trials as being useful in the treatment of a whole range of conditions. a clinical trial for parkinson's disease is currently in stage two. today, you put electrodes into a parkinson's patient�*s brain and it helps them

control their motor symptoms. rather than putting electrodes into the brain, we're able to put a tiny dose of a very specific gene that changes the neurotransmitters in just one site in the brain. and it re—circuits the motor signals and allows the patients to move normally. so it's an innovative way of using gene therapy, which involves local delivery of the gene that makes a signal and then changes how the brain processes in order to help with the symptoms of disease. it's important to note that this kind of gene therapy isn't yet a solution for everyone, and it does still have limitations. challenges still remain in manufacturing the genetic material and reducing side effects. and, of course, there may be

significant implications around cost. but the work being done by meiragtx and other medical companies has the potential, at least, to change lives. that was paul. one of my best friends has been blind since birth and i'm always in awe about how seemingly effortlessly he can navigate the world. his spatial memory is just incredible, and it's only when he's in a new location that he needs a little bit more assistance. but over the past few years, improved technology has really made a difference for how some blind and partially sighted people can get a better feel of the world around them. yeah, and we've been to meet one man who's been using some very smart specs. good dog. straight on. i'm stuart beveridge. i am visually impaired, totally blind, and have been since birth. good boy. the only sight i have is actually light and dark. so i know when a light�*s

on in a room and when it's not. and i know when the sun's out. everything is done by touch. i have no concept, really, of what anything looks like at all. ron is my second guide dog. i've had ron for nine years now, and ron basically helps me in my everyday life. he's with me 24/7 and it's great because, as well as being — 0k, he is a working dog — but he's actually my best friend as well, and we do almost everything together. so we're almost joined at the hip. independence — the dictionary defines it as the ability to live your life without being helped or influenced by others. introducing the new ai—powered smartglasses by envision. i've tried many different devices for visually impaired

people over the years, but in 2020 i bought these. the software runs on the google glass enterprise edition 2, which has a number of aids for visually impaired people. there's a speaker, there's a microphone, and they're controlled, really, by a touch pad, as well. artificial voice: connecting with robert beveridge. - hello. hi, dad. listen, i've mixed up my tins again. can you see it? yeah, i can see it. if you just — a wee touch higher. when i'm out and about, i use a feature called describe scene occasionally and itjust describes what's around me after taking a picture with the camera. so if my guide dog ron stops for any reason and i'm not sure why he's stopped, ijust take a picture with the glasses and it describes my surroundings. artificial voice: a person

holding a camera. - glasses beep. scan text. sighted people can just take things in at a glance. whereas that sense is totally beyond me. artificial voice: in terms i of team selection, i have got really good options right now and i was delighted | that we were able to sign ryan kent at the start. of this month. find object. find people. explore. my other four senses are having to work extra hard to compensate for the sight loss. it's the independence. i now don't need to rely so much on sighted assistance. i'm actually doing a lot more on my own. now time for a look at this week's tech news. twitter says it has stopped enforcing its policy on misleading information about coronavirus. a notice announcing the change appeared on the company's website but it hasn't commented

on why the change is happening. other policies on misinformation still do appear to be in place. controversial measures which would have forced big technology platforms to take down legal but harmful material have been axed from the uk's 0nline safety bill. critics claimed it posed a risk to free speech, but campaigners have accused the uk government of watering down the proposed laws. it was the government saying, yes, this is legal, you can say it to one another, but you can type it online and it is going to get banned by social media companies. that is a shocking place for us to end up. sony is aiming to bring motion capture tech for making avatars into our own homes through its latest release. available next year, these little devices are strapped parts of the body, allowing users to bring characters to life on the screen. it is going to be sold injapan first. and london's underground may be known for its network

of trains, but now farmers are taking over some of the space to house vertical farms. the technology has been installed in former world war ii air—raid shelters to grow food sustainably, and it means food can go from farm to plate much faster. phones, phones, phones. they're such a constant in most people's lives now — it's difficult to imagine being without them. and coming up to christmas, a new handset is on lots of lists to santa. but also on those wish lists, some people might like to add "better 56". here in glasgow, people really like their 5g. according to uswitch, across the country, about one third of us now have a 5g—enabled phone, but here, it's closer to half. however, the same study found that one in six people think the tech behind 5g has been... she tuts. ..a bit overhyped. i don't think it's great. i don't notice the difference. ijust don't think it's great. no.

sometimes if you're texting and that, like, it won't go through or, like, for example, i was trying to get hold of my mum but i couldn't get her. so i wasjust like, "oh, ijust need to wait." and most places have wi—fi now, as well. yeah. like, you just connect to the wi—fi if you're in a restaurant or shops. my provider- in sweden offers 56. so, like, there i have 56, but here i don't. i predominantly find most places are ag. yeah. and then the odd places, you get 5g, but not very much. when it works well, it's good. when it doesn't, it doesn't. if it doesn't work, it doesn't work at all. 36 barely works sometimes. my girlfriend, she's i got 56 on her phone, but there doesn't seem to be much of a difference - in comparison to my ag. so if it came to the price - was a big difference between 56 and 46, then i probably| would just stay with 46. most of the phones are going that way anyway. you know, the latest iphone comes 56 as standard. so, yeah, i think i will be going for a 56. yeah, definitely. i actually work with, like, social media and influencers. so, for me, it's probably needed now, as well, just that kind of more...

the extra boost. yeah. let me show you what i mean. we've got two phones here — one is running on 56, one is running on 4g. now, this is really unscientific — they're different phones and they're also on different networks. but what you can see is that, actually, the 4g is a lot faster in this particular spot than the 56. this isn't exactly helping 56 phones fly off the shelves in the shops. they don't really know how you sell that to a customer, cos, quite frankly, if someone walks into a shop, they're most likely to say, "i want an iphone," or, "i want a samsung galaxy phone" and it's got to have great battery, a nice screen, a nice camera. 56 is kind of left out there. so, is there enough 56 infrastructure in place? in many countries, we see the experience, things like speed, about five to six times faster on 56 compared with 46. at the top is typically korea, where average 56 speeds are well over 400 megabits now. the uk government's hoping that mobile internet can help plug gaps in hard—to—reach areas where it's difficult to install cables.

i asked 0fcom, the government's communication regulator, for an interview about 56 in the uk. it declined. instead, i spoke to mobile uk, the trade association for the main mobile network operators, and i asked why the 56 roll—out wasn't further ahead after all this time. well, in that three years, we've had covid, which did impact on our ability to deploy that network. equally, the government has taken decisions to take certain vendors out of the market, which has meant we've actually had to reconfigure networks, look for new vendors. and the government's own research suggested that cost an additional £2 billion and added a year to our deployment schedules. this centre at the university of surrey started working on 56 ten years ago. this is the team's message to people frustrated with 56. just be patient, because there are very interesting applications that's coming up. they're all gradually become 56 compliant,

all of these systems, and then we will see a huge amount of coverage. i'm not very good at driving this thing. they've been using this robot to demonstrate the difference in latency between 4 and 56. it was perfectly controlled using 56, but it all went a bit wrong. but look what happens the minute we switch it to 4g. heavy thud. ooh, god, that is really heavy! you are right! it's clear that the roll—out of 56 has had some problems — not only here in the uk but also around the world. mobile networks have got quite a lot of ground to cover before they can bring it to everyone but they all insist that they're on track. one of the biggest factors driving consumer uncertainty around 56 is price. like so many things, the cost—of—living crisis is making us think carefully about what we spend our money on. although 56 providers

don't charge extra money for the service, it can eat up your data at a much faster rate each month. while the telecoms companies are betting on the billions they've ploughed into 56 paying off, they won't be relying on their sales this christmas to bring it all home just yet. that was zoe. now, one in eight women will be diagnosed with breast cancer at some point in their lifetime, making it the most common cancer for women worldwide. and whilst five—year survival rates have massively risen, thanks to better treatment and better screening, there is always the fear that, after you've had treatment to get rid of it, the cancer could come back. mmm. but one company is using al to try and better predict recurrence rates, so that people get treatments that are more suited to them. gustave roussy in paris is one of the world's leading cancer centres, often treating rare or complex tumours, as well as trialling some of the latest diagnostics and therapies.

personalising a cancer plan can be a really tough challenge, especially when it comes to the balance between providing enough treatment to avoid recurrence, and over—treating and the impact that that can have on a patient�*s health. and here, what's known as a non—interventional clinical trial is using artificial intelligence to assess the risk of recurrence. but because this is a study where long—term data is vital, the patients aren't told of the al's findings, to avoid the risk of life—or—death treatment decisions being based on them. french and us tech company 0wkin has partnered with the hospital to digitise over 1,500 tissue samples from women with breast cancer. this will provide some of the information needed to classify patients between having a high, intermediate or low risk of their disease returning within five years.

we know that for this type of cancer, the prognosis is good, with more than 90% of patients be able to get rid of the disease. in these patients, we want to avoid as much as possible an over—treatment that can be considered in itself as a new disease, so that's why you want to identify the patients that are not likely to relapse. 0wkin aims to complement everyday clinical practice with its tool. it can already be used in european hospitals, and the company is working on an updated version for those in the uk. what does an oncologist, who's used to having difficult conversations with patients, make of it, though? what we do now in terms of assessment of the risk is to have the standard i information, the standard clinical and histopathological information — that are - the tumour size, the lymph node involvement, the tumour grade, |

the tumour histology, the i expression of the oestrogen. the ai is going to provide - a more comprehensive analysis of all this data and it's - going to complement these genomic tests, this genomic analysis. i pathologists here aim to digitise and analyse another 400 slides over the next two years as part of the prospective validation stage of the relapse risk study. to look at the patient�*s breast cancer tissue, the pathologist places the sample onto a slide, and then embeds it in paraffin wax. this helps both preserve it and keep it in place. then, instead of it being put under a microscope, it's taken from the lab to a digital scanner, where a very high—resolution image of the cancerous cells is created. digitising slides isn't a new concept, but running an ai model to determine the risk of cancer returning is. no—one involved is suggesting

that this will replace the importance of doctors, but the hope is that more detailed data and analysis could enhance theirjudgement calls. here you have the breast cancer that you can see. the thing that is white is fat tissue and the cancer, it's all this stuff which is kind of purple. so, i can see all these details on the slide and check that the quality is good. and once i say, "ok, the quality is good," i send it to a server, where the ai is performed. and then, after that, i receive a report performed by the ai, telling me what is the relapse risk associated to this slide. and, after that, i can check the report and also see if, for me, the relapse risk is in accordance with what i saw on the microscope.

by the nature of it, it needs to collect a lot of data, and that means it's going to take a long time before it can actually be properly useful. i think that a follow—up - of five years for the patients included in this trial - could be informative enough to understand the - performance of the ai. within minutes of the slides being scanned, the ai will have made its assessment of the risk. in this case, its low. but when it's at a more moderate level, the dilemma really exists. sometimes there is a grey zone and some cases are borderline, so maybe not completely good prognosis, maybe not completely bad, and in this grey zone, we need other tools to decide for the prognosis of the patient. and here, we have the information that you receive

after the al's looked at the data? exactly. so here, for example, i can see that the ai risk is low. it's a low—risk patient. some cancer experts see real promise in al tools like 0wkin�*s, but there are still challenges. it's really exciting - work and it's showing where the future i lies for pathology. while it might not be absolutely there yetl for clinical deployment, - i think it's a really important first step and something that we should be encouraged by. the real benefit to patients of these is patient safety . in the first instance, _ to make sure that we don't miss cancers, that we can detect small areas of cancer that i could otherwise be missed| by pathologists, but i think we're progressing along a journey and really, . without the digitisation i process, we won't be able to develop the rich algorithms that we need for patient - benefit and also, it would be - more challenging to deploy them into clinical practice.

so whilst it could be years before these systems are rolled out across cancer care, trials like this one help create the set—ups and data sets that could eventually revolutionise the way we treat patients. wow, how interesting! yes. and 0wkin�*s ai that you saw there is also being used for some research here at the crick institute. in conjunction with the royal marsden hospital, it's being used to look at the evolution of kidney tumours. how brilliant! mmm. well, i hope that you've enjoyed our look at the latest medical tech and techniques. thanks for watching, and we'll see you soon. bye— bye.

hello there. it's been cold enough alreadyjust recently, but i'm afraid to say it's going to get colder still over the coming few days. with some clear skies overnight, we're looking at widespread, quite sharp frosts. there is the threat of some icy conditions as well, particularly where we see some wintry showers. now, this colder air is coming our way because the wind is a northerly and that northerly wind sweeping down across the whole of the country now because it's a northerly northern scotland's really in the firing line. this is where we're going to see most of snow could be ten centimetres over the hills in the next few days. there is a risk of ice and snow elsewhere. but for many inland parts of england in particular, it's a very low chance of snow. it's going to be cold everywhere with a widespread frost, maybe not quite so cold early on wednesday in east anglia in the south east, because we've got a bit more cloud here that'll take a while to break up. otherwise many places will be dry and see some sunshine. it's areas exposed to that northerly wind that will see some wintry showers. parts of northern ireland down, some of these irish sea coast down the north sea coast

and more especially northern scotland. seeing that snow, those are the temperatures in the afternoon, 3 to five degrees, a little bit colder than it was on tuesday, windy in northern scotland and down those north sea coast. so it's these areas in particular that will feel quite a lot colder in that wind. and the snow continues as we head overnight and could get a bit heavier across northern scotland. we may see some mixture of rain, sleet and snow coming in to northern ireland as well. and that's coming into cold air, of course, another widespread frost with clear skies across many parts of the country. and we've got this cold air, this northerly air right the way across the country. now, the winds are not particularly strong. so where you have those clearer skies, that's why we're seeing such low temperatures at night. it could be icy for a while across northern ireland may turn a bit icy across western parts of wales with some wetter weather beginning to move in here. still some showers grazing those north sea coasts, more snow to come across northern scotland, although the snow may ease off later in the day. if anything, it may well be

a bit colder on thursday. typical temperatures only around two or three degrees. it doesn't get an awful lot better than that on friday either. another widespread, quite sharp frost. you can see we've got a few wintry showers, mainly around coastal areas. so in many, actually, the winds will be light. there'll be some sunshine around, but temperatures may be no better than around two to only five celsius. goodbye.

welcome to bbc news. i'm monika plaha. our top stories: morocco on the march. they beat former winners spain to reach the world cup quarter finals i am so happy! i am really happy! my boss back home in london, can you please give me another extension to the semifinal, please? i love you too. you are the best! they face portugal in the quarter finals. the team beat switzerland— with ronaldo left out of the starting line—up. we will have the latest on georgia as polls close in the crucial run—off election for the senate. a special report from haiti where an outbreak of cholera is adding to the misery of people in one of the world's most dangerous countries.

a jury in new york finds the trump organization guilty